Four outcomes, one table.
How markets price an election.

1. The chart

On the weekend before the vote, ZeroHedge shared a table on X titled “JPMorgan’s Midterm Scenario Map: What It Means For Stocks”. It gives four possible outcomes for Congress, the odds of each from Bloomberg prediction-market data as of November 1, 2026, and a stock-market reaction for each credited to JPMorgan Market Intel. We could not open the JPMorgan note itself or a web page carrying this exact table, so every number in it is the table’s own claim, labeled reported.

The table also lists political takeaways for each outcome. We leave those out and keep to the economics. Labels on the table such as “wave” are its own words; here the outcomes are named plainly, in the same order.

2. What “66%” means

A prediction market sells contracts that pay $1 if something happens and nothing if it doesn’t. If one trades at 66 cents, the crowd of traders is, in effect, putting the chance at about 66%. It is a price, not a poll. Anyone who thinks the true chance is higher can buy, and anyone who thinks it is lower can sell, and the price settles where those views balance.

Three things to keep in mind when you read one:

• It moves. A ZeroHedge write-up in early October said JPMorgan’s grid then gave “a 62% chance Dems sweep both the House and Senate”, up from “around 52% two weeks earler” (sic). By November 1 the table above says 66% reported.
• It can lean. The same write-up quotes Goldman Sachs: “Challenger sweep odds have typically faded about six points between now and Election Eve.” A price is a forecast, and forecasts drift reported.
• Check it against a second source. Saxo, writing with data as of October 29, said: “On 29 October, prediction markets saw Democrats winning both chambers as the most likely result, at roughly two in three.” That lines up with the table’s 66%.

One more check you can do yourself: the four odds in the table add to exactly 100% (66 + 26 + 7 + 1). Raw market prices often add to a little more than 100%, because traders pay a spread and fees. A clean table rescales them so they sum to one.

3. Why deficits move long-term yields

A long-term bond yield, like the 10-year or 30-year Treasury, is roughly two things added together: what investors expect short-term interest rates to average over those years, plus extra pay for locking money up that long. That extra pay is called the term premium.

When the government is expected to borrow more, it has to sell more bonds. To buy more of them, investors usually want more pay, so the term premium, and the yield, tend to rise. Expected borrowing cuts work the other way. That is the logic in the table’s yield column: it pencils in lower long-term yields where it expects less deficit spending, and higher yields where it expects deficit spending to stay high.

Saxo’s chief investment strategist, Charu Chanana, made the same link in describing a Republican sweep: “deficit and inflation worries could keep yields high.” Different firms, same mechanism. Whether either is right about the size of any move is exactly what the results will test.

4. Why markets price gridlock

When one party holds the White House and the other holds part of Congress, fewer big laws pass. For investors that means fewer surprises in taxes, spending and regulation, and predictability has value. Markets tend to reward it.

ZeroHedge’s October write-up summarized JPMorgan’s view this way: “Democrats will not win the two-thirds majorities needed to override vetoes, so a wave produces gridlock.” It also gave the history JPMorgan cited: “Since 1950, gridlock has coincided with S&P 500 gains of about 20%” over a two-year Congress, “versus 18% under single-party control” reported.

Two habits of a careful economist apply here. First, notice how small that gap is: two percentage points over two years. Second, notice the word “coincided”. A few dozen two-year stretches since 1950 is a small sample, and many other things happened in those years. History like this is a starting point for a question, not proof.

5. Do the arithmetic

A table of odds and outcomes invites one calculation: the probability-weighted expected move. Take the middle of each range, multiply by its odds, and add.

Using the low end of every range instead gives about +0.04%; using the high end gives about +0.69%. This is our arithmetic on the table’s numbers estimate.

Three lessons fall out of it:

1. Surprise moves prices, not news. The most likely outcome has the smallest reaction, because prices already lean toward it. The least likely outcomes carry the biggest moves. What matters is the gap between what happens and what was expected.
2. Find the puzzle. If all four outcomes were already fully in the price, the weighted reaction should be close to zero. Here it is slightly positive. One common reason: the end of uncertainty is worth something by itself. Saxo notes that “Since 1990, the S&P 500 has typically moved less than 1% on the day after a midterm, close to what is priced now, and the VIX usually fell.” Another possibility is simpler: the odds and the reaction ranges come from different sources and were never built to fit together. Holding both explanations open is part of the method.
3. Check the scale. Saxo also reported that “the options that expire on Wednesday price a move of about 1.3% in either direction by Wednesday’s close”, adding: “That covers the next four trading days, not only the vote.” Every range in the table sits inside that band. By the options market’s own pricing, the election is one event among several that week, not the whole story.

6. What to watch after the results

1. Which outcome happened, and when. Close races can take days to call, so control of one chamber may not be known on Wednesday.
2. Long-term yields. Did the 10-year and 30-year Treasury yields move the way the table’s yield column said?
3. The S&P 500. Was its move from Tuesday’s close to Wednesday’s close inside or outside the table’s range for the outcome that happened?
4. The VIX. Did the market’s fear gauge fall, as Saxo says it usually has after midterms?
5. The odds themselves. If a 7% outcome happens, that alone doesn’t make the market wrong. Things priced at 7% should happen about 7 times in 100. One result can’t grade a probability; many forecasts over many events can.

7. What happened update, Nov 4 markets, Nov 6

As of Wednesday morning, November 4, 2026, major outlets project that Democrats won control of both the House and the Senate. CBS News: “CBS News projects that Democrats will win control of the House and Senate in the 2026 midterm elections.” Al Jazeera, citing the Associated Press tally: “Democrats have won 228 seats in the House of Representatives so far, according to the Associated Press's tally, and are leading in 34 others,” and “In the Senate, the upper chamber of Congress, Democrats have won 54 seats and are leading in another, while the Republicans have won 44 seats and are leading in one.” reported

Not every race is settled. CBS News listed two Senate races still uncalled early Wednesday (Alaska and Mississippi) and said about 75 House races had yet to be determined. Final seat counts will move; control of each chamber is what the table scored, and that is what has been projected.

Score the table. The outcome that happened is the first row, priced at 66%. Its Brier score is the one worked in advance above: (1 - 0.66)² + 0.26² + 0.07² + 0.01² = 0.188. For scale, a forecaster who gave all four outcomes an even 25% would score (0.75)² + 3 × (0.25)² = 0.75, and a perfect forecaster 0. On this one event, the market odds did far better than a coin-flip spread. estimate

Two cautions, the same ones from section 6. One result can’t prove the odds were right: a 66% outcome should still fail about one time in three. And the odds were not a fixed number: they ran from about 52% to 62% to 66% over the month, so how you score depends on when you wrote yours down. That is why the assignment asked for odds before the results.

The market side update, Nov 6

On Wednesday, November 4, stocks rose far more than the table expected. TheStreet’s close: “S&P 500: 7,688.62, +119.70 (+1.58%)” and “Dow Jones: 53,356.12, +1,569.40 (+3.03%)”. The Motley Fool reported the same moves, with the Dow up 3.03% in what its headline called the “Biggest Post-Midterm Rally in 44 Years.” reported

Inside or outside the range? Outside. For the outcome that happened, the table’s S&P 500 range was -0.25% to +0.50%. The actual move, +1.58%, was more than three times the top of that range. It was also above the highest range anywhere in the table (+1.25%, for a Republican sweep) and above the roughly 1.3% that Saxo said options were pricing for the whole four days through Wednesday. estimate

Yields. The table’s call for this outcome was a “slight decline” in long-term yields. The Motley Fool reported that “the 10-year Treasury yield fell approximately 6 basis points to 4.89%.” That is the direction the table said. We did not find a 30-year yield or a Wednesday VIX figure in the sources we read, so we leave both open rather than guess. reported

Why so big? Two explanations fit, and the method is to hold both. One is surprise: the result may have been stronger than the odds implied, or the market valued the end of uncertainty more than the table did. The other is that the election was not the only news. TheStreet wrote that “Energy stocks saw the steepest haircuts as oil benchmarks fell today on the midterm results,” and other coverage tied the day’s gains partly to the fall in oil prices. A single day’s move mixes every story at once, which is why the table’s ranges, and our scoring of them, are rough tools.

What this scores. The odds side of the table did well (0.188 against 0.75). The reaction side missed: the right direction for yields, but stocks moved well outside the range priced for this outcome. Good odds and a good reaction estimate are separate skills, and this table shows one without the other on this one event.

Sources

• ZeroHedge on X, “JPMorgan’s Midterm Scenario Map: What It Means For Stocks”, an image crediting ZeroHedge, JPMorgan Market Intel and Bloomberg prediction markets, odds as of November 1, 2026. Read as an image on November 2, 2026. reported
• ZeroHedge, “Wall Street’s Midterm Map Just Flipped - Here’s What Goldman And JPMorgan See”, early October 2026 (no date shown on the page). Read November 2, 2026. reported
• Saxo, Koen Hoorelbeke, “US midterms: the options market expects a quiet Wednesday. What if it is wrong?”, market data as of October 29, 2026 (no publish date shown). Read November 2, 2026. reported
• CBS News, “Live election results for 2026 midterms as Democrats decisively win control of House and Senate”, updated November 4, 2026, 5:34 AM EST. Read November 4, 2026. reported
• Al Jazeera, “Democrats win control of House and Senate: Key US midterms takeaways”, published November 4, 2026, with seat tallies credited to the Associated Press. Read November 4, 2026. reported
• TheStreet, Rob Lenihan and Noah Weidner, “Stock Market Today (Nov. 4, 2026)”, updated November 4, 2026, 4:20 PM EST. Read November 6, 2026. reported
• The Motley Fool, RJ Fulton, “Stock Market Today, Nov. 4: Dow Surges 3% in Biggest Post-Midterm Rally in 44 Years”, November 4, 2026, 4:50 PM EST. Read November 6, 2026. reported

Drafted by Claude (Anthropic) on November 2, 2026, America/Los_Angeles, at Michael Hoydich’s request, from the three sources above, each read that day. The table’s numbers are the table’s own claims; the weighted move and Brier scores are our arithmetic. Section 7 was added on November 4, 2026 from the CBS News and Al Jazeera pages above, each read that day; the market numbers were added on November 6, 2026 from the pages by TheStreet and The Motley Fool listed above, each read that day. Report a correction.

More from the University of El Segundo →