The Lever Board

Which Western levers actually move the priced odds of a Ukraine-favorable outcome, what does each cost — and why does the market price nearly all of them as un-pulled?

TON618 Capital Research · Polymarket marks as of 24 August 2026

The market prices peace talks by year-end at 51% and a signed peace deal at 14.5% — a 1-in-3.5 conversion — and prices every deliberate Western lever that could raise that conversion below 10%, while pricing an involuntary NATO–Russia clash at 26.5%. The revealed expectation is drift: pressure rising, no lever pulled to completion, the war grinding through the Kostiantynivka–Druzhkivka line (both ~52%) without breaking the Kramatorsk–Sloviansk belt (~28–33%) and without conquering Donetsk (2.9%). The settlement weight, meanwhile, has been moving toward terms Ukraine would call unfavorable: across the archived vintages since August 8, the deal-without-cession residual — the priced weight on a settlement without territorial concession — has compressed from 10 points to 4, driven by the deal leg falling (19.5% → 14.5% since 8/8) while the cession leg held its ground (10.0% → 10.5%, round-tripping through 6.5% on 8/22).

Fact vs. opinion, stated up front: every percentage in this note is a market price, not a house forecast. Where we interpret (“the market believes X”), the inference is stated and testable. The house takes no side in this war; “Ukraine-favorable outcome” is used throughout as a defined, priceable basket, not an endorsement.

Defining the outcome so it can be priced. A Ukraine-favorable outcome = a ceasefire or settlement without the concession legs — no territorial cession (10.5%), no armed-forces cap (7.0%), no forced NATO renunciation (7.0%) — plus a real security guarantee (US 8.5% / any European state 7.0%). On today’s board the deal-without-cession residual is roughly 4 points (deal 14.5% minus cede 10.5%, treating cession as a subset of deals) — down from ~7 points two days ago.

Alongside the charted gates: on Path A, talks by 9/30 print 11%, ceasefire by 10/31 8.5%, and a Putin–Zelensky meeting 6.5% (93.5% “never meet”); a European guarantee 7.0%. On Path B, Russia entering Dobropillia prints 88.5% (near-done), Druzhkivka 51.5%, Sloviansk-enter 27.5% (capture 7.0%), Lyman 34.5%, Kupiansk 27.5%; the terms rungs — give up the rest of Donbas 4.2%, cap the armed forces 7.0%, forswear NATO 7.0%. Three reads fall out of the ladder:

READ 1Talks ≠ deal. 51% for talks, 22.5% for a ceasefire, 14.5% for a signed deal: the market’s central case is a meeting that does not close. Conditional on talks occurring, a deal converts at roughly 28%. Path A’s gate 1 is open; gate 2 is where the probability mass dies.

READ 2The settlement, as priced, is drifting toward Path B’s terms — but the driver is the deal leg falling, not the cession leg rising. Across the archived vintages the deal-without-cession residual has compressed from 10.0 points (8/8: deal 19.5% − cede 9.5%) to 7.0 (8/22) to 4.0 (8/24). The three-week view disciplines the 48-hour one: the cession leg’s two-day jump (6.5% → 10.5%, on a $712k book — not thin) is a round-trip back to its early-August level (10.0% on 8/2), while the durable moves are the deal leg falling 19.5% → 14.5% and the year-end ceasefire leg falling 35% → 22.5%. Either way the arithmetic lands in the same place: of the settlements the market still prices, the share involving territorial cession has risen from roughly half (8/8) to roughly three-quarters (8/24). Our pre-registered reversal tell — cession outrunning the deal leg — fired on the 48-hour window; the three-week series says the squeeze is coming from the deal side, and the cession signal stays unconfirmed until the 8/31 strip resolves.

READ 3The battlefield feeds the table; it does not close the war. Cession at 10.5% against Donetsk-conquest at 2.9% means the market prices any concession as negotiated, not imposed — Russia cannot take the land it would be ceded, on these odds. Path B, as priced, is a pressure lever on Path A’s gate 4, not an independent road to an ending.

Three reads, one question left standing: who can move these gates? Seven things Western allies could do; the gate each moves; what each costs; and the market’s price on it being pulled.

The exhibit’s one-line finding: every lever the market can price directly sits below 10%, and the only contract above 25% among the seven levers’ direct contracts is the one nobody chooses. (The wider ladder has higher legs — talks 51%, territorial rungs to 88.5% — but none of them is a Western choice.) The market’s revealed expectation is drift — sustain, pressure at the margin, pull nothing to completion. The prices show what the crowd expects the West to do, not what the West prefers.

The exception that proves the mechanism: the one lever pair demonstrably in use is 2/3 — coercion at the margin. Its signature sits in the term structure — ceasefire mass parked in Q4 (1.6% by 8/31 → 8.5% → 22.5%) — read with care: cumulative contracts price higher at longer horizons mechanically, so the signal is not the upward slope itself but the pairing of a 51% talks leg with a 22.5% year-end ceasefire leg, and the fact that the 12/31 ceasefire leg has fallen from 35% on 8/8 while the talks leg sits at 51% (steady across the two vintages archived for it). Pressure is working on the meeting, not yet on the signing — and the 8/22→8/24 cession move suggests the market thinks the signing, if forced, now tilts toward Moscow’s terms.

Expected-value framing for any portfolio is out of scope by design: this note prices outcomes. The probabilities partition (77.5 + 8 + 10.5 + 4 = 100) under the stated assumption that cession occurs only inside a signed deal; the risk section takes that assumption back up.

1. Thin books on the territorial mid-ladder. Kramatorsk $53k, Sloviansk-enter $63k, Kupiansk-capture $22k. Thin rungs are flagged ±5pts and never carry a conclusion alone. The legs the conclusions actually rest on are deep: ceasefire-12/31 $2.2M, deal $2.6M, cede $712k, clash-12/31 $1.6M, NATO membership $1.2M.
2. Polymarket resolution risk. “Enter/capture” contracts resolve on adjudicated map sources; disputes at the margin are possible. Date-bounded Kalshi twins, where they exist, are the cross-check.
3. The 48-hour cession move could be noise or a single large trader. It is flagged as an early tell, not a confirmed regime change; confirmation requires the pattern holding through the 8/31 strip resolution.
4. Attention bias. Polymarket’s book skews to Western participants; a systematically pro-Kyiv or pro-drift bias in the crowd cannot be excluded and is not correctable from inside the data.
5. Model risk in the residual arithmetic. “Deal minus cede = coerced-settlement weight” assumes cession only occurs inside a deal; a unilateral-freeze-with-de-facto-cession outcome would break that identity.

A conclusion is owed, and the board’s own arithmetic supplies one — read as prices, not policy advice; the house holds no side, and every claim below is conditional on the objective being the Ukraine-favorable basket as this note defined it.

Are the levers under-utilized? The board cannot answer that directly — it prices political will, not lever efficacy (the steelman’s second untestable). But it supports three precise statements. First, the only levers in visible use are the coercion pair (2/3), and their yield is flattening: enough pressure to convene a meeting (talks 51%) but not to close one (deal 14.5% — down net from 19.5% over three weeks, though up 1.0pt in the latest 48 hours). Second, every deliberate lever is priced below 10% while the involuntary outcome — a NATO–Russia clash — is priced at 26.5%: drift is not the risk-free path; it already carries the board’s single largest risk. Third, the 4-point Ukraine-favorable residual is, mechanically, a priced bet that the terminal levers stay holstered.

Which levers, ranked by probability moved per unit of escalation risk — on this board’s arithmetic:

1. Industrial co-production and air defense (levers 4/6) are the cheapest probability on the board. They move no peace leg directly, but they are what holds Path B’s gates shut — and the market already credits them (Donetsk conquest 2.9% against a 53% Kostiantynivka leg). Their binding risk is industrial (the component chain), not escalatory.
2. The security guarantee (lever 5) is the highest-value single pull. It is the terminal gate on both paths — the only lever whose pull would reprice every downstream leg at once, which is exactly why the reversal section names the guarantee legs the leading indicator. Its cost is a contingent liability, not a present escalation; its risk is that it cannot be un-pulled.
3. The coercion pair (levers 2/3) is working at diminishing returns. Demonstrated bite, measurable tells — and each further increment spends against the 26.5% clash budget while the deal leg keeps falling.
4. The money (lever 1) has a proven de-risked pattern. The ERA structure showed a low-risk tranche of the lever can be pulled — profits, not principal — without triggering the seizure objections. The remaining ~$300B principal is the untested rest of that lever, and the steelman’s seizure-cost case stands against it unrebutted here.
5. Boots (lever 7) is dominated on this board. Near lever 5’s terminal effect at a strictly higher escalation cost; the market prices it accordingly (5.2%).

The one lesson that does not depend on which side of the steelman is right: as priced, the West is currently paying escalation risk without buying probability on any peace leg — a 26.5% clash contract alongside a falling deal leg is the board’s worst quadrant. Moving out of it in either direction — pulling a terminal lever, or genuinely de-escalating — buys something the current posture does not. Drift buys nothing and still pays the premium — unless restraint itself is what holds the clash leg at 26.5% rather than higher (the steelman’s first untestable), in which case drift is buying the ceiling; that is the one reading this lesson cannot rule out. Whether the exit should run through lever 5 or through restraint is the choice this note prices but does not make.

The honest residual — the case in which no lever suffices. Every ranking above assumes Moscow’s binding constraint is strategic: costs weighed against aims. If it is instead regime-internal — if the war economy, mobilization politics, and elite structure now require the war itself — then no terms clear, the tolerate-versus-fight comparison never binds, and the grind is not a failure to settle but the objective. The market’s 77.5% no-ceasefire base case, sitting beside a deal leg that has fallen net for three weeks and a talks leg steady at 51% in our archive, is arguably that reading, priced. On that view the levers’ realistic product is not a settlement at all — it is a cheaper, more stable stalemate: Path B’s gates held shut at declining cost, escalation risk contained, Ukraine’s de facto security built without anyone’s signature. That is a lesser goal than peace, and this note names it as the one the base case actually pays out.

Everything above is a reading of prices. What follows is the house’s conditional speculation beyond what prices alone establish — opinion, labeled as opinion, still conditional on the Ukraine-favorable basket being the objective, and still owing the steelman its due: the restraint reading of this same board would sequence nothing and negotiate now. With that stated, the arithmetic implies an order, and the order has a logic: buy probability before spending risk, and pull reversible levers before irreversible ones.

1. First — the no-regret tier: scale levers 4 and 6, funded through lever 1’s proven tranche. Co-production and air defense spend no escalation budget, are scalable and reversible, and hold Path B’s gates shut — which is what keeps the battlefield feeding the table instead of replacing it. The ERA pattern (profits, not principal) already showed how to fund this without triggering the seizure objections. Why first: every later lever works better from behind a holding line, and none of this forecloses a single future choice.
2. Second — re-point lever 5: the West’s guarantee to Ukraine activates at signing, not before it. To be precise about who gets what: the guarantee is to Kyiv, from the West — Russia gets nothing except certainty about what signing buys. The mechanism is two-sided, and honesty requires both halves. For Kyiv, this is what makes a deal signable at all: an unguaranteed freeze is a rearmament pause, and that constraint — not only Moscow’s reluctance — is a candidate cause of the 1-in-3.5 talks-to-deal conversion. For Moscow, cutting the other way, a guarantee-at-signing raises the price of signing: it locks in a Western-committed Ukraine, which is exactly what Russia says it fights to prevent. On its own, lever 5 re-pointed this way could therefore lower Moscow’s willingness to close. It works only paired with the no-regret tier: with levers 4/6 compounding, not signing means facing an increasingly fortified, Western-industrialized Ukraine anyway — on a slower fuse, at continuing cost. The choice put to Moscow is a guaranteed Ukraine behind a frozen line now, or an effectively backed Ukraine later from a worse position — waiting stops paying in either branch, and that pairing, not the guarantee alone, is the outlast-killer. Structure: European tranche (7.0% leg) before US (8.5%), bilateral before treaty-grade; committed-if-signed is a different risk object than committed-now, which converts lever 5’s worst property (irreversibility) into its point.
3. Hold, don’t increment — the coercion pair (2/3). Maintain, don’t add — the readout’s arithmetic already made this case, and the marginal coercion dollar is better spent on the no-regret tier. Coercion’s remaining job is not to win — it is to keep the price of waiting positive while levers 4/6 compound.
4. Keep holstered — boots (7) and lever 1’s principal. Boots stays dominated by the guarantee at every point on this board. The ~$300B principal stays where the steelman left it: the seizure case is unrebutted here, and the profits tranche is not yet exhausted.

Who signs what — the paradox and its resolution, with the market’s confirming bids. An Article-5-style guarantee is the thing this war was launched to prevent; no pressure short of collapse makes Moscow sign one, because signing converts every cost paid since 2022 into strategic defeat regardless of territory held. The market prices that refusal as its deepest conviction on the board: NATO membership before 2027 at 3.3% on a $1.2M book. The resolution is that Russia’s signature is required on the ceasefire line — not on the guarantee. Korea 1953 is the template: the armistice was signed with the adversary; the US–ROK Mutual Defense Treaty was signed with the ally, separately, just over two months later — the other side never consented to the treaty, only to the line.

In parallel papers, Moscow signs a freeze while the West signs the guarantee with Kyiv; Putin does not concede it — he tolerates it, as Moscow tolerated armed West Germany and armed South Korea. The face-saving decomposition even has its own leg on the board: Ukraine formally forswearing NATO membership is priced at 7.0% while the bilateral guarantee legs sit at 8.5% / 7.0% — de jure neutrality alongside de facto commitment, each separately priced, is a package the market treats as more likely than membership by a factor of two or more. And this is why the sequence leads with levers 4/6: they are the levers that need no Russian consent at all. A fortified, Western-industrialized Ukraine is the de facto guarantee, Russia holds no veto over it, and every month of grind builds the thing Moscow would be tolerating at signing anyway — the choice becomes acknowledging it from behind a frozen line now, or facing more of it from a worse position later. The template’s limits are stated plainly: in 1953 the treaty’s credibility was garrison-backed, and the adversary at the table was not the nuclear principal; here the de facto industrial guarantee substitutes for the garrison — an untested substitution, and the restraint case’s strongest objection to this architecture.

The accelerant clause: this sequence assumes time is neutral. The reversal section says when it is not — if the cession leg keeps outrunning the deal leg through 9/30, the settlement being priced is drifting toward Moscow’s terms, and the guarantee-at-signing move stops being a closing instrument and becomes the floor under Ukraine’s negotiating position. In that branch it moves from second to first.

How we would be wrong: if the steelman’s restraint case is right, step 2 reprices the clash leg before the deal leg and the sequence has spent the board’s scarcest budget on its least certain mechanism. That is the bet, stated plainly — and tripwire 4 (clash clearing 35%) is where this speculation concedes.