Traders Had Lula Winning Brazil's First Round. Then the Count Started.

Contracts on Lula topping Sunday's vote traded near 70 cents an hour before polls closed and were worth almost nothing by midnight. Flávio Bolsonaro now trades around 82 cents to win the Oct. 25 runoff, a price the vote math only partly supports.

Prediction markets spent Sunday afternoon confident that Luiz Inácio Lula da Silva would finish first in Brazil's presidential election. Within two hours of the polls closing, that confidence was gone.

With 99.96% of ballots counted, Flávio Bolsonaro took 47.04% of valid votes, or 56,098,129, against 45.15%, or 53,845,563, for Lula. Because neither reached a majority, the two will meet in a runoff on Oct. 25. The winner takes office on Jan. 5, 2027.

How fast the price moved

On Polymarket, a contract paying $1 if Lula won the most votes in the first round traded at 69.5 cents at 3 p.m. Eastern and 66.5 cents at 4 p.m., the hour polls closed in Brasília. By 5 p.m. it had collapsed to 12.5 cents. It reached 3.05 cents at 6 p.m. and 0.05 cents by 11 p.m.

Kalshi's equivalent contract followed the same path. It changed hands between 70 and 71 cents until shortly after 4 p.m., fell to 59 cents within half an hour and then to 56 cents, and was trading between 1.3 and 8 cents by about 5:45 p.m. Overnight it settled at fractions of a cent.

The miss was not unique to traders. The final major poll before the vote showed Lula ahead by 42% to 38%. The actual result was a 1.89-point lead for Bolsonaro, a swing of nearly 6 points on the margin.

The runoff price

Contracts on the overall winner moved the other way. Polymarket's Bolsonaro contract rose from 63.65 cents at 4 p.m. to a peak of 85.15 cents an hour later, then eased to 81.85 cents by early Monday. Kalshi's contract traded at 81 to 82 cents at the same point.

The question is whether 82 cents reflects the arithmetic of the runoff or the momentum of election night.

The first-round numbers give a framework. Candidates eliminated on Sunday collected about 7.8% of valid votes, roughly 9.3 million ballots. Ignoring changes in turnout, Bolsonaro needs about 38% of those voters to reach a majority. Lula needs about 62%. That is a meaningful edge for Bolsonaro, but it is not an 82-to-18 edge on arithmetic alone; it assumes transfer patterns favor him or that Lula's own voters turn out in smaller numbers in three weeks.

One election night is also not a track record. The first-round collapse shows how quickly these markets reprice on hard information. It says nothing about whether the opening runoff price is well calibrated.

What to watch: The first trading session for the real and Brazilian stocks after the vote, and the first runoff polls that measure where voters for eliminated candidates are heading. A transfer split near 50-50 would argue that the Bolsonaro contract is running ahead of the math.

d