Kalshi Politics — Mispricing Audit (Anchor, NO-side structural)

Report date 2026-10-19 · strategy=Anchor · audit_version=anchor · data as of 2026-10-19T16:40Z · capital $2,500 · horizon 45 days · universe: targets.json (120 series, generated 2026-10-19T16:36Z)

1. Summary

Picks1

Logged rejects26

Capital deployed$124.955.0% of capital

Cash$2,375.0595.0%

One leg: NO on “Hegseth out as Secretary of Defense before Dec 1, 2026” at 85¢, half size (thin-favorites band), sized to 147 contracts (half-size cap binds; depth cap 161). The slate was thin by construction: of 885 open markets in the 120 allowlisted series, only 14 close inside 45 days with ≥500 contracts of 24h volume and a tradeable quote, and nine of those are transit-count or executive-order-count ladders that the Anchor rules exclude. The Bank of Canada October ladder is live and liquid but is a forecast of a decision's content at a sub-band price on both legs. The Anthropic IPO, US–Iran deal and Tesla/SpaceX merger rungs are all structurally NO but priced at or beyond the 90¢ deep-favorite line. Cash at 97% is the expected result of an efficient week.

Honesty note on edge. The one pick's written edge is small (pre-market 12% → final 14% YES, versus a 15/18 book). Anchor sizes flat and EV-agnostic because the Sep 2026 record showed status-quo NO picks in the 85–89¢ band winning 85% against ~83% implied; the pick is deployed on shape, band and mechanism, not on a claimed large dislocation. If you require ≥3¢ of written edge, skip it and hold 100% cash.

2. Universe and funnel (Stage 0)

Step	Markets	Note
Open markets in allowlisted series	885	Kalshi public API, status=open, 120 series
Closing beyond 45 days	−823	out of horizon
volume24h < 500	−34	live number
YES ask ≤ 5¢ / YES bid ≥ 95¢	−9	logged as deep-favorite band where notable
Expires inside 7 days	−5	Suez Oct 12–18 ladder, logged
Raw candidates	14
Continuous-number markets (Suez transit counts ×7, Mamdani EO count)	−8	logged
Stage-2 candidates researched	6	Hegseth Dec 1 · BoC hold · BoC hike · Anthropic IPO Dec 1 · US–Iran Dec 1 · Tesla/SpaceX Dec 1
Picks	1

No candidate carried volume24h > 5000 except the Bank of Canada legs (37k / 16k), which were specialist-dense and are rejected on band and shape anyway. Tier 1 series yielded zero tradeable candidates inside the horizon (every departure/ouster, legal-action and confirmation rung either closes in 2027 or trades ≤5¢).

3. Pick

BUY NOmechanism: status-quoconviction: MEDIUMarchetype: multi-news-othercluster: hegseth-sod-exit

`KXHEGSETHOUT-26AUG-26DEC01` — Will Pete Hegseth be out as Secretary of Defense before Dec 1, 2026?

YES condition: for YES to pay, Hegseth must actually vacate the office (announced intent does not count; temporary leave does not count) before Dec 1, 2026 04:59Z, resolved by Kalshi on the vacancy. This is a new action by a person or institution, so the market is Anchor-shaped.

Thesis. Status quo: Hegseth is still in office on Dec 1 — every dated pressure vehicle (House returns Nov 9 with a tabling vote promised; defense funding deadline Dec 11; NDAA) lands at or after the cutoff, Trump's own words remain supportive, and the rules pay only on an actual vacancy, not an announced exit.

Probability. 12% pre-market → 14% final for YES (book 15/18, mid 16.5). Fair NO ≈ 86¢ vs 85¢ paid; edge 1¢ vs the ask, 2.5¢ vs mid; pre-market framing would be 88¢ fair. Resolver-risk deduction 2 pts (death clause → last fair price; 'vacate and re-occupy' clause) already inside the 14%.

Vehicle / calendar / already-happened (primary sources).

Vehicle: Rep. Massie's eight-article privileged impeachment resolution (introduced Sep 15). Speaker Johnson adjourned the House a day early on Sep 16 to avoid it and said it "will immediately be tabled, of course … We will do it when we come back"; the clock resumes when the House returns Nov 9 (NBC News, Sep 16; Fox News). Even a successful impeachment does not vacate the office.
Calendar: Senate Democrats are floating Hegseth's ouster as the price of lame-duck cooperation on defense funding that expires Dec 11, the NDAA and the Hung Cao Navy-secretary waiver (Yahoo/The Hill, Oct 19). Every leverage point sits after this rung closes. The same piece quotes a GOP senator that expecting Trump to give up Hegseth is "a bridge too far".
President's own words: "I am extremely happy with the job that Pete Hegseth is doing" (Truth Social, Aug 6); "fantastic job" (Aug 31). On Oct 15 Trump overruled Hegseth's plan to livestream the Dec 3 Fort Hood execution; Pentagon spokesman Parnell: "The Secretary supports the President's decision, and they are completely aligned" (Axios, NBC, Washington Post, Oct 15). No Presidential statement since signals a change.
Structural: the Iran war is ongoing (~50,000 troops deployed; Trump said no strikes before Nov 3), there is no confirmed Army or Navy secretary and no vetted successor (Newsweek, Sep 3). Precedent-store lesson on this series: "an active war … is a strong structural NO for the sitting defence chief".
Already happened? No. The only firm "Trump is considering firing him after the midterms" claim is an unsourced X post by a commentator (AOL, Oct 16).

Precedents consulted: KXHEGSETHOUT-26APR-JUN01 (NO); KXHEGSETHOUT-26APR-JUL01 (NO); KXHEGSETHOUT-26APR-SEP01 (NO); KXHEGSETHOUT-26AUG-26OCT01 (NO, API-verified this run); KXTRYFIRECOOK-27JAN01-26OCT01 (NO) — 'price the calendar of the President's own process before the rhetoric'. Lessons leaned on: "Congressional resignation demands and IG findings have never resolved this archetype Yes … require an affirmative White House signal"; "price this series off the President's own words, not the volume of criticism".

Cleanest tail (new discrete event, acceptable): Trump fires Hegseth effective immediately in the days after Nov 3 (the Sessions Nov 7, 2018 pattern), or Hegseth resigns after a failed tabling vote on/after Nov 9. This needs a new Presidential act, not a continuation of the criticism trend. Priced at roughly 12 of the 14 points. A Rumsfeld-style announcement with departure on a successor's confirmation would not pay before Dec 1 under rules_secondary ("must have an actual departure date and vacate the role within the time period").

Entry context (Stage 2.5): YES 16/18 at signal (NO 82/84 nominal, but only 2 contracts sit at 16¢ so the effective NO fill is 85¢), spread 2–3¢; 483 YES bids at 15¢ and 1,000 at 14¢; volume24h 2,100; OI 87,542 (up from ~60k on Oct 14). 30-day history: 23–25¢ on Sep 20–23, 13–18¢ Oct 3–14, spike to 36¢ close / 62¢ intraday on Oct 15 (livestream overrule), 24¢ Oct 17, 17¢ Oct 19. 48h move −7¢ toward our side: explained (headline spike fading) and not news we lack. Sibling FEB01 rung 39/42 — the crowd concentrates departure risk in Dec–Jan, consistent with the thesis.

Sizing: Band 85–89¢ → half size = 5% of capital = $125 → 147 contracts; depth cap = 1/3 of the 485 YES bids a NO buy at 85¢ lifts (2 at 16¢ + 483 at 15¢, re-quoted 16:50Z) = 161 contracts. Final: 147 contracts ≈ $124.95 max loss (two contracts fill at 84¢).

4. Position sizes and limits

Ticker	Side	Limit	Contracts	Cost = max loss	Max win	% capital	Band / size rule
`KXHEGSETHOUT-26AUG-26DEC01`	NO	85¢	147	$124.95	$22.05	5.0%	85–89¢ thin favorite → half size ($125 cap, binding); depth cap 161
Cash				$2,375.05		95.0%	reserve ≥10% satisfied

5. EV for the record

Ticker	P(YES) pre	P(YES) final	Market mid	Fair NO	Paid	Edge ¢	EV/contract	EV %	EV $ (147)
`KXHEGSETHOUT-26AUG-26DEC01`	12%	14%	16.5¢	86¢	85¢	+1	+1.0¢	+1.2%	$+1.47

EV is recorded, not used for sizing (Anchor sizes flat). Win ≈ +$22.05 (86% model / ~83% market-implied); loss ≈ −$124.95.

6. Cluster table

Cluster	Deployed leg	Runner-ups / related rejects
hegseth-sod-exit	DEC01 NO	FEB01 (outside horizon, weaker status quo), NOV01 (YES ask ≤5¢)
boc-oct28-decision	—	H0 (yes-side thesis / below band), H25 (below band)
anthropic-ipo-timing	—	DEC01 (below band / yes-side), NOV01 (deep favorite)
us-iran-deal-2026	—	DEC (deep favorite), NOV (deep favorite)
tesla-spacex-merger	—	DEC01 (deep favorite), NOV01 (deep favorite)
suez-weekly-count / mamdani-eo-count	—	continuous-number markets

7. Risk profile

Single-name, single-event exposure: the whole book is one 147-contract leg; worst case is −$124.95 (5.0% of capital).
Named tail: immediate post-midterm firing or resignation (Nov 4–30). Monitor Trump's own statements after Nov 3 and the House floor on Nov 9–10 (tabling vote). A Senate-Democrat "price" for Dec 11 funding would land after the cutoff.
Resolver: Kalshi on a vacancy; death clause settles at last fair price; "vacate then re-occupy" settles on initial vacancy. Deduction 2 pts.
Liquidity: 87k OI, 3¢ spread; exit before expiry is feasible but not assumed — this is a hold-to-settlement position.
Conflict of interest disclosed: the auditing model is built by Anthropic. The Anthropic IPO rung was researched from public sources only and is rejected on band and side rules; no position is taken either way.

8. Execution notes

Buy NO at a limit of 85¢ (lifts the 2 YES bids at 16¢ and then the 483 at 15¢). Do not chase above 85¢: 86¢ would still be in band but the fill would come from the 1,000-contract 14¢ YES bid wall only if it is still there; re-check the book.
Slippage budget ≤3¢; expected 0 at 147 contracts against 485 visible.
Close 2026-12-01T04:59Z; expected expiration (settlement) 2029-01-01 per the API, but the market closes early on a qualifying departure. Settlement after the Dec 1 close is at Kalshi's discretion on the facts; past rungs finalized within ~7 days.
Do not add the FEB01 rung as a "hedge" or second leg — it is outside the horizon and spans the dated vehicles.

9. Logged rejects (26)

Market	Archetype	Side @ price	P(YES)	Reason
`KXCBDECISIONCANADA-26OCT-H0` Bank of Canada maintains rate at Oct 28, 2026 meeting	scheduled-process	YES @ 59¢	55%	yes-side thesis: hold-YES forecasts the decision's content, not a process step (precedent store KXCBDECISIONCANADA-26SEP-H0 says exactly this); also below band at 59¢. Sep CPI 3.1% y/y but ex-gasoline 2.2% (StatCan, Oct 19) vs CORRA swaps ~65% hike (Bloomberg, Oct 19) — genuinely contested, no Anchor shape.
`KXCBDECISIONCANADA-26OCT-H25` Bank of Canada hikes 25bps at Oct 28, 2026 meeting	scheduled-process	NO @ 58¢	45%	below band: NO costs 58–59¢; swaps-implied hike odds (~65%) are above Kalshi's 42%, so the NO has negative edge vs the swaps curve (precedent lesson: 'NO on the hike rung has no edge at Kalshi prices').
`KXIPOANTHROPIC-DATE-26DEC01` Anthropic officially announces an IPO before Dec 1, 2026	corporate-event	NO @ 35¢	60%	below band (NO at 34–36¢); the YES side fails the side rule — no Anthropic S-1 is on EDGAR as of Oct 19 (EDGAR company search: 'No matching companies'; full-text S-1 search returns only SpaceX amendments), no ticker assigned, and the 'Oct 23 filing / Nov 9 roadshow' timeline is an analyst inference, not an official calendar → yes-side thesis. Conflict-of-interest note: the auditing model is built by Anthropic; public sources only were used.
`KXUSAIRANAGREEMENT-27-26DEC` US agrees to a new Iran nuclear deal before Dec 1, 2026	scheduled-process	NO @ 94¢	4%	deep-favorite band: NO costs 92–94¢. Trump said Oct 8 he no longer wants a deal (CNBC); Iran reviewing US proposals with preconditions (Al Jazeera Oct 12); 06/17/26 clarification already ruled the MoU non-qualifying. Structural NO but ≥90¢.
`KXCOMPANYACTIONMERGER-27-26DEC01` Tesla/SpaceX definitive merger agreement announced before Dec 1, 2026	corporate-event	NO @ 92¢	8%	deep-favorite band: NO costs 91–92¢. No agreement, special committee or SEC disclosure of talks as of Oct 12 (BNN Bloomberg); Barclays 'possible this year' is analyst speculation. Structural NO but ≥90¢.
`KXHEGSETHOUT-26APR-27FEB01` Pete Hegseth out as Secretary of Defense before Feb 1, 2027	multi-news-other	NO @ 60¢	35%	cluster runner-up / outside horizon: closes Feb 1, 2027 (>45d). Also the window contains the dated pressure vehicles (Nov 9 impeachment clock, Dec 11 funding, lame-duck NDAA), so the status-quo case is materially weaker than the Dec 1 rung.
`KXMAMDANIEO-26OCT24-T0` Mamdani signs >0 executive orders Oct 18–24, 2026	confirmation/enact	NO @ 76¢		continuous-number market: weekly executive-order count ladder that slipped past the classifier; not researched.
`KXSUEZWEEKLY-26OCT25-T270` Suez Canal transit calls Oct 19–25 above 270	legal-action	YES @ 92¢		continuous-number market: PortWatch weekly transit-count ladder (misfiled archetype); not researched.
`KXSUEZWEEKLY-26OCT25-T280` Suez Canal transit calls Oct 19–25 above 280	legal-action	YES @ 80¢		continuous-number market: PortWatch weekly transit-count ladder (misfiled archetype); not researched.
`KXSUEZWEEKLY-26OCT25-T290` Suez Canal transit calls Oct 19–25 above 290	legal-action	YES @ 52¢		continuous-number market: PortWatch weekly transit-count ladder (misfiled archetype); not researched.
`KXSUEZWEEKLY-26OCT25-T300` Suez Canal transit calls Oct 19–25 above 300	legal-action	NO @ 67¢		continuous-number market: PortWatch weekly transit-count ladder (misfiled archetype); not researched.
`KXSUEZWEEKLY-26OCT25-T310` Suez Canal transit calls Oct 19–25 above 310	legal-action	NO @ 76¢		continuous-number market: PortWatch weekly transit-count ladder (misfiled archetype); not researched.
`KXSUEZWEEKLY-26OCT25-T320` Suez Canal transit calls Oct 19–25 above 320	legal-action	NO @ 86¢		continuous-number market: PortWatch weekly transit-count ladder (misfiled archetype); not researched.
`KXSUEZWEEKLY-26OCT25-T330` Suez Canal transit calls Oct 19–25 above 330	legal-action	NO @ 94¢		continuous-number market: PortWatch weekly transit-count ladder (misfiled archetype); not researched.
`KXSUEZWEEKLY-26OCT18-T320` Suez Canal transit calls Oct 12–18 above 320	legal-action	NO @ 89¢		inside 7d (expires Oct 20); also a continuous-number market.
`KXSUEZWEEKLY-26OCT18-T310` Suez Canal transit calls Oct 12–18 above 310	legal-action	YES @ 51¢		inside 7d (expires Oct 20); also a continuous-number market.
`KXSUEZWEEKLY-26OCT18-T300` Suez Canal transit calls Oct 12–18 above 300	legal-action	YES @ 69¢		inside 7d (expires Oct 20); also a continuous-number market.
`KXSUEZWEEKLY-26OCT18-T290` Suez Canal transit calls Oct 12–18 above 290	legal-action	YES @ 84¢		inside 7d (expires Oct 20); also a continuous-number market.
`KXSUEZWEEKLY-26OCT18-T280` Suez Canal transit calls Oct 12–18 above 280	legal-action	YES @ 94¢		inside 7d (expires Oct 20); also a continuous-number market.
`KXDIESELEXPORTBAN-26` US diesel export ban in effect before Nov 3, 2026	multi-news-other	NO @ 96¢		already an active pick (EXISTING_PICKS); also YES ask ≤5¢.
`KXRETIREMM-26` Mitch McConnell resigns before election day 2026	departure/ouster	NO @ 99¢		deep-favorite band: YES ask ≤5¢ (mechanical cut), NO would cost ≥95¢.
`KXUSAIRANAGREEMENT-27-26NOV` US–Iran nuclear deal before Nov 1, 2026	scheduled-process	NO @ 98¢		deep-favorite band: YES ask ≤5¢ (mechanical cut), NO would cost ≥95¢.
`KXIPOANTHROPIC-DATE-26NOV01` Anthropic IPO confirmed before Nov 1, 2026	corporate-event	NO @ 99¢		deep-favorite band: YES ask ≤5¢ (mechanical cut), NO would cost ≥95¢.
`KXCOMPANYACTIONMERGER-27-26NOV01` Tesla/SpaceX merger agreement before Nov 1, 2026	corporate-event	NO @ 95¢		deep-favorite band: YES ask ≤5¢ (mechanical cut), NO would cost ≥95¢.
`KXHEGSETHOUT-26AUG-26NOV01` Pete Hegseth out before Nov 1, 2026	multi-news-other	NO @ 95¢		deep-favorite band: YES ask ≤5¢ (mechanical cut), NO would cost ≥95¢.
`KXIPOSTARLINK-26DEC01` Starlink IPO before Dec 1, 2026	corporate-event	NO @ 98¢		deep-favorite band: YES ask ≤5¢ (mechanical cut), NO would cost ≥95¢.

10. Precedents consulted and writeback

Consulted: KXHEGSETHOUT (3 rows + OCT01 verified via API), KXTRYFIRECOOK (2), KXFOMCDISSENTCOUNT (2), KXCBDECISIONCANADA (13), KXIPOANTHROPIC (3) + KXIPO-26-SPACEX, KXSUEZWEEKLY (2), KXHORMUZ* (archetype rows). Nothing in the store for KXMAMDANIEO, KXCOMPANYACTIONMERGER or KXUSAIRANAGREEMENT; those were derived cold from the live rules text and clarifications.

Written to /tmp/out/precedents.json: one row for KXHEGSETHOUT-26AUG-26OCT01 (settled NO, API-verified; vehicle calendar for the lame duck; headline-spike-and-fade pattern on the DEC01 rung).

Precedent quoted on the pick (KXHEGSETHOUT-26APR-SEP01): "A cabinet-level resignation by a subordinate plus a same-party senator calling for the firing is still not an affirmative White House signal … Price this series off the President's own words, not the volume of criticism."

11. Method notes

Stage-0 universe from targets.json only; live quotes from the Kalshi public API (*_dollars/*_fp fields); price history from the candlesticks endpoint; rules text and clarifications from /markets/{ticker}; precedents from resolution_precedents via the read-only query API (schema fetched first). Probabilities were committed before opening the book for every Stage-2 candidate. Sources cited inline: Kalshi API; NBC News (Sep 16, Oct 15); Fox News; Axios (Oct 15); Washington Post (Oct 15); Yahoo/The Hill (Oct 19); AOL (Oct 16); Newsweek (Sep 3); Kurdistan24/Truth Social (Aug 6); Statistics Canada The Daily (Oct 19); Bloomberg via Yahoo Finance Canada (Oct 19); BlueGamma; FinanceFeeds (Oct 12); SEC EDGAR company and full-text search (Oct 19); CNBC (Oct 8); Al Jazeera (Oct 12); BNN Bloomberg (Oct 19).