South Carolina Senate Special 2026 — Runoff

Three days out, and the book has closed twenty points. Darline Graham was 87.5% five days ago. She is 67.5% now, against Ralph Norman at 33%—and the money did not follow the favourite. Of $786,947 traded in the last 24 hours, $417,330 went through Norman's contract and $369,617 through Graham's. The trailing candidate is outtrading the leading one on the way up.

Open interest has followed. $2,112,359 sits on Graham and $2,011,343 on Norman—within 5% of each other on a book priced two-to-one. Somebody is very sure this is closer than the headline.

The margin book agrees with the money rather than the winner book. Its two deepest positions are Norman by 3-6% ($41,476) and Norman by 0-3% ($39,698)—both narrow Norman wins, and together more open interest than any Graham margin.

One more thing worth knowing about this book: eight eliminated candidates are still carrying money. Mark Sanford's finalized leg holds $778,377 and Russell Fry's holds $508,867—more open interest on two people who cannot win than most live books on the exchange carry in total.

The runoff on Kalshi, next to the vote-share ladder that disagrees with it.

The winner book prices Graham at 67.5%. Her own vote ladder implies 52.2% of the vote. Those two numbers are not incompatible—a 52/48 race genuinely does favour the leader—but two-to-one is a strong price for a five-point margin three days out, and the margin book's open interest is positioned for something tighter still.

(Manifold lists this race but has not been marked in meaningful size; it was dropped from this page rather than shown as a stale mirror beside a $787k book.)

The two contracts, side by side. Graham at 67.5% and Norman at 33%—which sums to 100.5%, one of the tightest two-way books on the exchange.

Norman took more volume than Graham in the last 24 hours ($417,330 to $369,617) and holds within 5% of her open interest. On a book that prices him at a third, that is the single most interesting fact on this page.

The margin and turnout complex, which is where the runoff's real disagreement lives.

Margin ($65,100): the modal outcome is Graham by 6-9% at 13.5%, but the two deepest open-interest legs are both narrow Norman wins—$41,476 on Norman 3-6% and $39,698 on Norman 0-3%. The most-traded single leg is Graham by 24% or more at 5.5%, which took $13,912. The book is barbelled: a Norman squeaker or a Graham blowout, with the middle thinly held.

Turnout ($70,276, $114,420 of open interest): implies roughly 308,000 votes. Its deepest position is $25,351 on Above 200k at 96%.

The two vote-share ladders are internally consistent, which is rare. Graham's implies 52.2% and Norman's implies 47.3%—summing to 99.5%. That is a 4.9-point margin, considerably tighter than the 67.5/33 winner book implies and much closer to what the margin book's open interest is positioned for.

The county map, and round one as the record. Kalshi lists which counties Graham takes: Lexington at 93.5% and Horry at 90.5% are near-locks, Charleston at 57.5% and Oconee at 53% are coin flips, and the upstate is Norman's—Anderson at 25%, Pickens at 23%, Newberry at 40.5%.

That is the whole race in one book. Norman's path is running up the upstate margin far enough to cover Lexington and Horry, and the county contracts price that as unlikely but not remotely dead.

The seat in November, where the market is almost bored. Kalshi prices Republican hold at 89.2% against Annie Andrews at 11%.

Note the open interest, though: the Democratic contract carries $230,737 against the Republican's $221,475—more money on the 11% side than the 89% side. That is the only sign anywhere in the South Carolina complex that somebody thinks a bruising runoff could put this seat in play.

• Whether Norman's flow is right and the price is wrong. He is outtrading the favourite, holds nearly matching open interest, and owns the two deepest margin legs. Either that is a coordinated position or the winner book is slow.
• The 4.9-point implied margin against a 67.5/33 winner book. Those two are hard to reconcile. A 52.2/47.3 race should not price the favourite at two-to-one.
• Turnout above 308,000. Runoff turnout is the classic upset variable, and the book's deepest position is a 96% rung—meaning almost nobody is positioned for a collapse.
• Anderson and Pickens. If Norman wins, it happens there. Both currently price him at roughly three-to-one to carry them.
• The $1.29M on eliminated candidates. Sanford and Fry's finalized legs will unwind at settlement. Watch whether that changes the shape of the live book when it does.
• The general-election open-interest inversion. More money on the Democrat at 11% than the Republican at 89%. Worth revisiting the week after the runoff.