Kalshi Politics — Mispricing Audit Keel v4 · archetype-targeted

Run date 2026-08-05 · Universe: targets.json (120 pre-classified series, 70 Tier-1 / 50 Tier-2) · Horizon 45 days · Capital $2,500

Verdict: 0 picks — 100% cash $2,500 reserve

This is a correctly-priced week, and v4 says to say so. The 45-day window is dominated by one Senate floor week (Aug 3–7, recess Aug 10 → after Labor Day) and one hot news cycle (Trump vs. Pirro). Both are being traded by liquid, specialist-dense books that repriced faster than any thesis we could form from public sources: the Blanche confirmation rung moved 12¢ → 85¢ in 72 hours as the committee standoff repaired; the Pirro Sep-1 rung round-tripped 46¢ → 71¢ → 30¢ around the Oval Office confrontation. Every Stage-2 survivor failed the edge, band, or 48h-move gate and is logged below with a committed probability, so the reject log shadow-tracks this run's judgment.

Stage-0 universe
53
open target-series markets ≤ Sep 19
Mechanical cuts
49
42 dead books · 6 tail · 1 deep-fav
Stage-2 diligenced
4
+1 cluster runner-up logged
Picks
0
cash is a position

Market environment

Rejected after full diligence 5 logged + 1 active-pick exclusion

KXAGCONF-26JUN05-SEP01 — Blanche confirmed as AG before Sep 1 confirmation/enact

Considered: YES @ 86¢ effective · cluster blanche-ag-confirmation · book 84 / 85, 18.6k vol24h, 41.1k OI · Reason: no edge vs. efficient book

Prob: 88% pre-market → 84% final (blended toward a very liquid book). The bull case is real and mostly mechanical: committee 12–10 after the holdouts' named demands were met, cloture filed this morning, recess as forcing function, floor votes scheduled Fri/Sat. But the residual 15¢ is not free: Collins is a declared NO with a margin of one; Murkowski and Cassidy are uncommitted; a withdrawal resolves NO immediately (dated resolver clarification in rules_secondary); and any slip past Saturday is a mechanical NO because the Senate does not return before Sep 1. Deductions: floor defection −8, schedule collapse into the CR fight −4, withdrawal −2. Fair ≈ 84 vs. 86¢ effective entry (only 13 contracts visible at 85; real depth at 86) → edge ≤ 0 in the thin-favorite band. The book already repriced 12¢ → 85¢ on the exact thesis in 72h.

Precedents consulted: KXAGCONF-26-JUN15 (NO — price the clock from formal transmittal, 5–6 week floor); KXAGCONF-26-JUL01 (NO — "first announced pick" keys to the current vacancy); KXBILLS-HOUS (YES — enactment mechanics beat headlines); KXRECNCH-26-JUN06 (NO — upstream passage does not compress the other chamber's floor schedule).

KXPIRROOUT-26AUG-26SEP01 — Pirro announces departure before Sep 1 multi-news-other

Considered: NO @ 70–71¢ effective · cluster pirro-out · book 29 / 30, 14.5k vol24h, 18.7k OI · Reason: 48h move already captured / public-info-only

Prob (YES): 25% pre-market → 27% final. The departure-archetype precedent corpus leans NO: not fired at the confrontation, no successor floated, principal fighting to keep the office ($25k retention bonuses), and "considering removal" sourcing has never resolved this archetype YES. But the market got there first — the rung spiked 46¢ → 71¢ on meeting day and collapsed to 30¢ within 24 hours of the "not fired" reporting. Selling the spike was the trade on Aug 4; at 70–71¢ effective today the residual edge is 2–3¢ of pure archetype base-rate against a specialist-dense book, in a market where a single Truth Social post resolves YES (announcement counts regardless of effective date). That is a public-info-only thesis inside slippage.

Precedents consulted: KXKASHOUT-26APR-JUN01 (NO — "matter of time" sourcing is not a vacancy; a principal publicly fighting back holds); KXLUTNICKOUT-26JUL01 (NO — absent a named successor, scandal doesn't move this archetype); KXLUTNICKOUT-26JUN01 (sell the pre-event spike, don't hold through); KXTRUMPADMINLEAVE-26DEC31-TGAB (YES — announcement/last-day mechanics).

KXPIRROOUT-26AUG-26AUG06 — Pirro out before Aug 6 multi-news-other

Book 5 / 7, 1.8k vol24h · cluster pirro-out · Reason: entry band (YES 7¢ < 40¢ floor; NO 93–95¢ ≥ 90¢)

Prob (YES): 5%. Requires an announcement within hours. Nothing pending as of the Aug 5 afternoon news cycle. Both sides fall outside the v4 entry bands; tails are 0-for-8 lifetime.

KXPIRROOUT-26AUG-26AUG07 — Pirro out before Aug 7 multi-news-other

Book 4 / 5, 24.6k vol24h · cluster pirro-out · Reason: mechanical tail cut (YES ask ≤ 5¢) + cluster runner-up

Prob (YES): 7%. Same thesis as the Aug 6 rung with one more day of Trump-whim exposure. Logged so the cluster shadow-tracks.

KXJUDGECOUNT-26AUG-1 — ≥2 federal judges confirmed in August confirmation/enact

Book 7 / 10, 605 vol24h · cluster judges-aug-count · Reason: entry band (YES 10¢ < 40¢ floor; NO 90–93¢ deep-favorite)

Prob (YES): 15% pre-market → 13% final. Verified zero eligible confirmations Aug 1–5; the July blitz (Flowers, Martin, Pozos, Traynor) landed Jul 22–24 and does not count; recess Aug 10 → Sep 7 makes this week the entire August window; and Thune's Aug 5 cloture filings included no judicial nominees (judges are excluded from en bloc batching, and CIT/Tax Court confirmations don't count under rules_secondary). The only YES path is two judicial cloture petitions by Thursday amid a CR fight — possible, not probable. Slight positive EV at 10¢ but hard-banned; logged to shadow-track the <40¢ rule.

Precedents consulted: no series precedents in store (noted); archetype analogues KXRECNCH-26-JUN06 (count scheduled session days, not calendar days) and KXAGCONF-26-JUN01 (a rung inside a mechanical dead window is a structural NO).

KXCLARITYVOTE-26JUL-AUG08 — Senate recorded vote on CLARITY Act before Aug 8 existing pick

Reason: already an active pick — excluded from new picks by hard rule.

Observed in passing this run: Thune publicly committed to a CLARITY Act vote before the recess — consistent with the live YES position.

Stage-0 mechanical cut log (49 of 53)

Cut ruleMarketsNotable examples
Live volume24h < 500 (zero-volume phantom-edge rule)42 All KXSENATECONFIRM legs, KXLAKECONF, KXKASHOUT-SEP01, KXLEAVEWALZ, KXHEGSETH monthlies (some had quotes but dead books), IPO series, most Banxico/BoC strikes, KXIMPEACH-SEP01 (421), KXTRUMPTHUNEOUT (371)
YES ask ≤ 5¢ (tail)6 KXPIRROOUT-AUG07, KXIPOANTHROPIC-SEP01, KXLEAVEPOWELLGOV-SEP01, KXHEGSETHOUT-SEP01, KXALITOOUT-SEP01, KXUSAIRANAGREEMENT-26SEP
YES bid ≥ 95¢ (deep favorite)1 KXCBDECISIONMEXICO-HOLD (96 bid)

The August recess explains the dead books: with the Senate about to leave and no scheduled catalysts, 42 of 53 target-series markets traded fewer than 500 contracts in 24h. The 98-of-136 zero-volume lesson from the resolved record makes these hard cuts, not judgment calls.

Portfolio & risk

ClusterBest legDecisionCapital
blanche-ag-confirmationKXAGCONF-26JUN05-SEP01 YESReject — no edge at 86¢$0
pirro-outKXPIRROOUT-26AUG-26SEP01 NOReject — 48h move captured$0
judges-aug-countKXJUDGECOUNT-26AUG-1 YESReject — band (<40¢)$0
Cash$2,500 (100%)

Conviction exposure

TierPositionsCost% of capital
HIGH0$00%
MEDIUM0$00%
Cash reserve$2,500100%

Execution notes

Method note (v4)

Universe restricted to the pre-classified target allowlist; probabilities committed before studying the book, then blended; Stage-1 precedent retrieval before opinion formation; one leg per cluster; entry bands 40–84¢ core / 85–89¢ half-size / hard rejects outside; sizing EV-agnostic (HIGH 15% · MEDIUM 7% of capital, halved in the thin-favorite band), depth-capped at ⅓ of visible contracts; every non-pick survivor logged with side, price, and probability so the record can grade the rules as well as the picks. Data: mirror DB via query API (schema-first), Kalshi public API (*_fp fields, candlesticks endpoint), primary-source news (senate.gov daily press, uscourts.gov confirmation listing, NPR/WaPo/NBC/CNN/WTOP).