2026 Florida Governor Race

Byron Donalds is the Republican nominee for governor of Florida. Kalshi's nominee book settled Yes on Donalds at 99.9%, exactly as it had been priced for weeks, and on the way there it turned over $19,672,109 in twenty-four hours against $56,784,498 of open interest—four times yesterday's volume and, by a wide margin, the deepest single book this catalog has ever tracked.

Almost none of that money was on the winner. Polymarket's mirror is the cleaner read because it settled cleanly: James Fishback's leg took $3,459,813 of the $4,392,764 that traded there, against $574,522 on Donalds. Seventy-nine percent of the flow, on a leg priced at a tenth of a percent, that resolved No.

And the market got the margin wrong. Yesterday Donalds' vote-share ladder priced at least 50% at 92%. It now prices 1%—at least 45% is the last rung standing at 99%. Donalds won the nomination and landed between 45 and 50 percent of the vote, a full bucket below the book's central estimate.

Fishback, priced at 1.1% to win, cleared between 10 and 11 percent. His at least 10% rung was a literal coin flip at 50% yesterday; it resolved Yes by less than a point.

Florida Division of Elections certifies on September 1.

Both exchanges called the winner and neither was ever in doubt. Kalshi settled Donalds at 99.9% on $19.7M of twenty-four-hour volume; Polymarket at 99.9% on $4.4M. The pies below are the final book.

What the two exchanges disagreed about was what to list. Polymarket carried seven names and stopped. Kalshi carried nine plus eight satellite books on vote share, placement and exact finishing order—and those satellites are the only Florida contracts still trading today.

With the nominee settled, $3,843,406 of twenty-four-hour volume has rotated into Fishback's vote-percent ladder—the deepest live prediction market contract on this page and one of the deepest anywhere.

The two legs below are the ones that carried it, and both are instructive. At least 30% took $933,084 at a tenth of a percent, the single most-traded rung in the whole complex and one that was never plausible. At least 10% took $600,003 and is the rung that actually decided something, resolving Yes against a 50/50 price.

Read together: the deepest leg was the least informative, and the informative leg was third-deepest.

Kalshi lists Cumulative vote-share ladders on both the winner and the leading challenger. Overnight they resolved in opposite directions relative to how they were priced.

Donalds' ladder is the miss. At least 45% holds at 99%; at least 50% prices 1%. Twenty-four hours ago that same 50% rung was 92%, and 60% was 43.5%. The market was certain about the nominee and a full bucket wrong about his share—and its deepest legs, at 50% and 60%, are the two that broke.

Fishback's ladder is the near-miss. At least 10% prices 99.8%, at least 11% prices 1%. He landed in a one-point window. The rungs above—15%, 20%, 30%—all went to a tenth of a percent, and they are where $1.8M of the day's $3.8M traded.

Neither ladder settles for months: Fishback's book is dated January 1, Donalds' a full year out, even though both are effectively decided. Prices that read 99% and 1% here are a certified-result formality, not a live question.

Second place went to Jay Collins, and the money was on the other side. Collins' leg prices 99% now; Fishback's prices 1%—and Fishback's leg took $182,160 to Collins' $86,557, two to one on the wrong side of the outcome. That was the one genuinely live question in the book yesterday, and the flow called it wrong.

The exact-order book confirms it: Donalds–Collins–Fishback at 99%. It is the only Florida contract with a near-term settlement, dated September 1 on certification, which is why the Countdown above now points there.

Then November. The general-election book has Republicans at 78% and Democrats at 23%—still summing above 100, still surviving on $33,037 a day. On the Democratic side David Jolly won the nomination outright, settling Yes at 99.7%, so that pie below is now a record rather than a race. Florida's next tradeable question is a two-way general, and $169,342 of open interest is the whole of it.

• Where the $56.8M of open interest goes. The nominee book settled with more outstanding than it traded. Whether that liquidity rotates into the general, the placement ladders, or leaves Florida entirely is the single best read available on whether primary depth is event-specific or sticky.
• The gap between decided and settled. Fishback's ladder is dated January 1 and Donalds' August 2027, both pricing 99/1 on outcomes already known. Any residual trade in those books between now and certification is pure carry, not information.
• Certification on September 1. The exact-place book is the only Florida contract with a near-term date. Donalds–Collins–Fishback at 99% is a formality unless a recount changes an order.
• Whether the margin miss repeats. The book was right about the winner and a bucket wrong about the share, and its deepest legs were on the rungs that broke. The same structure is live right now in South Carolina's runoff and Ohio's special—worth checking whether vote-share ladders are systematically mispriced or Florida was one bad draw.
• The general's 101%. Republican 78% plus Democrat 23%. It persisted through the primary on thin flow; if the post-primary rotation lands here, it should close.