Middle East Escalation Tracker

Every leg of this stack has moved in the same direction over the past three weeks, and it is not toward escalation or resolution. It is toward neither.

The nuclear-deal odds collapsed. Kalshi's KXUSAIRANAGREEMENT priced a deal before March 2027 at 35% a month ago. It prices 11% now. Before the end of the Trump term: 53% then, 38% now. A twenty-four point and fifteen-point drop with no headline attached.

The Hormuz recovery keeps getting pushed out. Return to normal before January 2027 was 54% in July, 38.5% last month, and prices 29% today. Before July 2027 has gone 67% → 53.5% → 46%. Three consecutive markdowns of the same curve.

And the war book has not moved. Polymarket's US invades Iran before 2027 sits at 17%, where it has been for weeks—but its volume has fallen from roughly $180,000 a day to $75,301. Regime fall prices 7%; Manifold's mirror prices 4%, so the rare cross-exchange convergence this workspace used to note has opened back up to three points.

Put together: the market is pricing no American invasion, no negotiated settlement, and no reopening of the strait. That is a frozen conflict, and it is a genuinely unusual thing for an event contract complex to converge on—escalation books normally resolve toward one tail or the other.

Anchor: Polymarket will-the-us-invade-iran-before-2027 · $8,296,782 of open interest.

The concession ladder prices the intermediate steps between the status quo and the headline outcomes, and it has flattened toward zero.

Uranium surrender: August 31 at 0.9%, October 31 at 3%, December 31 at 7%—down from 8.5% a month ago. US custody of the stockpile: August 31 at 0.8%, September 30 at 2%, December 31 at 6%. Four rungs of each ladder—April, May, June, July—have already resolved No.

Leadership change has come down too: December 31 at 16% from 18.5%, and June 2027 at 25% from 32.5%. Note the August 22 rung, three days out, priced at 0.1% and carrying $3,227 of the day's volume—the second-most-traded leg in that book, on a proposition it prices at one in a thousand.

Kalshi's deal Multiple is where the collapse is cleanest. Before September 1 prices 1% and took $12,115, the most-traded leg in the book, while holding $917,446 of open interest. The most active contract in the nuclear-deal complex is a rung the market says will not happen in twelve days.

US intervention before 2027 at 17%, Iranian regime survival at 7%. The ratio still prices American action without collapse—strikes and pressure rather than a change of government.

What has changed is the money. The invasion book carries $8,296,782 of open interest against $75,301 of daily volume, and the regime book $4,553,052 against $3,639. Both are now overwhelmingly position rather than trade: a hundred-to-one and a thousand-to-one ratio. These are hedges being carried, not views being expressed.

The mirror has diverged again. Polymarket prices regime fall at 7%, Manifold at 4%. This workspace previously noted these as a rare convergence; they are three points apart now, on $3,639 and effectively nothing.

The physical books are the ones deteriorating. This week's transit Cumulative (8/17–8/23) prices above 30 at just 16%—it was 49.5% for last week's book—and above 25 at 50%. The modal week has fallen from roughly thirty transits to roughly twenty-five inside seven days.

Beside it, highest single-day August traffic: at least 30 prices 5%, at least 40 at 3%, at least 100 at 1%. No single day this month is expected to clear thirty ships.

And the normalization Multiple carries $9,302,285 of open interest—the largest in this stack—of which $1,954,444 sits on a before July 1, 2026 leg that already settled No and $1,378,089 on before September 1, 2026 priced at 1%. A third of the book's outstanding money is on rungs that are dead or nearly so.

• A frozen conflict is the base case now, and that is the finding. No invasion (17%), no deal (11% by March 2027), no reopening (29% by January). Three books that should trade against each other are all pricing "nothing resolves." Whichever one breaks first is the trade.
• The deal book fell twenty-four points on no news. 35% to 11% for a March 2027 agreement, and 53% to 38% for one before January 2029. Repricing that large without a headline usually means a participant with a view unwound or arrived—worth watching whether it retraces.
• Hormuz recovery has been marked down three times running. 54% → 38.5% → 29% for before-January-2027. The market is not waiting for normalization; it is serially discovering that normalization is further away. That is the single most consistent trend in this workspace.
• Volume left the war book, positions did not. $8.3M of open interest against $75k of daily volume on the invasion contract. A hundred-to-one ratio means this is a hedging instrument now, and hedges do not reprice on news the way views do.
• $3.3M stranded on dead Hormuz rungs. Nearly two million dollars on a leg that settled No in July, and $1.4M more on a leg priced at 1%. Any "open interest" figure quoted for this complex is measuring mostly that.
• The weekly transit book is the real-time tell. Modal week fell from thirty transits to twenty-five in seven days, and single-day max above thirty prices 5% for all of August. The political books are frozen; the physical one is still getting worse.
• Cross-exchange depth is asymmetric and the mirror has drifted. Polymarket carries the political volume, Kalshi the deal and Hormuz books, Manifold nothing. The regime-fall mirror is three points apart again—treat it as a prior, not a check.