Report date 2026-09-23 · strategy Anchor · audit_version anchor · focus category Politics · horizon 45 days · capital $2,500

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

0picks deployed

22logged rejects (7 reached Stage 2)

$0allocated · 100% cash

901open markets scanned in 117 allowlisted series

1. Verdict

Zero picks. Cash is the position this week.
Every Anchor-shaped market inside the 45-day horizon is already priced as a deep favorite (NO ≥ 90¢) or is a sub-60¢ macro coin flip. The two names within a cent of the 85–89¢ lane (Patel Nov 1, Tesla/SpaceX Nov 1) trade at 90¢ marketable and would need a resting order behind existing depth to get 88–89. Anchor does not bend the band for that.

Why nothing cleared. The Anchor record lives in NO entries at 60–89¢ on event markets. In the Politics allowlist right now, that price lane is empty inside the horizon: the departure ladders (Patel, Hegseth, Miller, Gallego, McConnell, Pelosi) have all decayed to YES ≤ 12¢, the Anthropic Nov 1 rung collapsed to 6–8¢ on the Sep 18 WSJ delay report, and the only two-sided market (Bank of Canada Oct 28) is a 44/56 hawkish repricing whose loss tail on either side is the trend continuing. The in-band NO candidates that do exist (Hegseth before 2027 at NO ~73¢, Patel before 2027 at NO ~83¢, Wiles, Lutnick, MJ rescheduling) all close Jan 1, 2027 — 100 days out, beyond this run's 45-day horizon — and are listed in section 6 for the next horizon window, not traded here.

2. Method (Anchor)

Universe: targets.json (Stage-0 allowlist, 120 rows / 117 unique series, min OI 5,000). 901 open markets pulled from the Kalshi public API; live volume_24h_fp used for the 500 floor.
Mechanical cuts: beyond horizon 810 · vol24 < 500: 820 · spread > 5¢: 351 · YES ask ≤ 5¢: 171 · close after expiration: 72 · inside 7d: 12 · YES bid ≥ 95: 6 (overlapping). 7 survivors reached Stage 1–2.
Order of operations: precedent rows → rules text + rules_secondary → YES-condition sentence → primary-source news → pre-market probability committed → book/candles opened → blend → resolver deductions → Stage-3 hard rules (side, continuous-number, continuation tail, rules-reading) → band → cluster → flat size.
Side rule: NO default; YES only on already-happened or process-in-motion. No YES thesis qualified this run.

3. Picks

None. No position-size, limit, EV or cluster tables apply. Reserve held: $2,500 (100%).

4. Stage-2 candidates (all rejected) — pick cards

departure/ouster NO side deep-favorite (≥90¢)

Kash Patel out as FBI Director before Nov 1, 2026?

KXKASHOUT-26JUL-26NOV01 · closes 2026-11-01

REJECT — deep-favorite band. NO is 90¢ marketable (YES bid 10, only 13 contracts) and 92–93¢ at size; status-quo NO is the right side but the band caps entry at 89¢

YES condition: For YES to pay, Kash Patel must actually vacate the FBI Director role (resignation, removal, term end — not a leave of absence or a 'considering replacing' report) before Nov 1, 2026, resolved by Kalshi on the actual departure date.
Thesis shape: Anchor-shaped (new action by a person/institution). Mechanism if traded: status-quo.
Probability (YES): Prob: 8% pre-market → 7% final. Market YES 10/12; NO effective entry 90¢.
Vehicle: No vehicle. No resignation letter, no named successor, no White House statement. The last dated primary-source event is the Sep 15, 2026 Senate Judiciary oversight hearing (CBS/WaPo/NBC live coverage) — a five-hour grilling that Patel finished in office, with Republicans on the panel not calling for his exit. Searches for a Sep 22–23 ouster story return only late-2025 'Trump weighing replacement / Andrew Bailey' pieces (New Republic, MS NOW) that never produced a departure.
Calendar: 39 days to the Nov 1 deadline. No confirmation process needed for an acting replacement, so a firing could land any day; but the record on this series (KXKASHOUT APR-JUN01, APR-JUL01 both NO at 67–77% priced) is that news-cycle pressure does not convert to a vacancy inside a 5–6 week window.
Already happened?: No. Patel is in office as of Sep 23 (testified Sep 15; no departure report since).
Precedents consulted: KXKASHOUT-26APR-JUL01 (NO): 'far rungs of monthly departure ladders are systematically overpriced… each 30-day window adds only the low base rate of an actual firing.' KXKASHOUT-26APR-JUN01 (NO): 'Anonymous it's-only-a-matter-of-time sourcing is not a resolution event; a principal publicly fighting back has not been told to go.' KXHEGSETHOUT-26APR-JUN01 (NO): 'require an affirmative White House signal before paying above base rate.'
Resolver profile & risk: Kalshi single-official read of an actual vacancy; rules_secondary requires 'an actual departure date and vacate the role within the time period'. Death → last fair price (–1 pt). Re-occupation settles on initial vacation (–0). Total deductions ≈ 1 pt.
Cleanest tail: Cleanest loss: Trump fires Patel or Patel resigns before Nov 1 — a new discrete event, not a continuation. Acceptable tail shape. Not the blocker.
Book / 48h move: Candles: 5–6¢ Sep 17–19 → 8–9¢ Sep 22 (5.3k vol) → 18¢ last print Sep 23 on a book with 13 contracts at the YES bid. 48h move is a thin-book narrative pop with no primary-source catalyst found; it argues the NO is cheap-ish at 90–93 but not that the band should bend. Entry context: YES bid 10 / ask 12, depth at touch 13 contracts, vol24h 4,651, OI 35,912.

corporate-event NO side deep-favorite (≥90¢)

Tesla/SpaceX definitive merger agreement before Nov 1, 2026?

KXCOMPANYACTIONMERGER-27-26NOV01 · closes 2026-11-01

REJECT — deep-favorite band. NO is 90¢ marketable (276 contracts at YES bid 10); resting at 89 sits behind 170 contracts already bid; no-vehicle NO is correct but outside the band

YES condition: For YES to pay, Tesla or SpaceX must officially announce (press release, SEC filing, verified account, or Musk's unambiguous confirmation of a signed agreement) a definitive, binding merger/acquisition agreement before Nov 1, 2026.
Thesis shape: Anchor-shaped (institutional action by a deadline). Mechanism if traded: no-vehicle.
Probability (YES): Prob: 6% pre-market → 7% final. Market YES 10/11; NO effective entry 90¢.
Vehicle: No vehicle. No 8-K, no board announcement, no proxy. Musk's most recent public comment was at the All-In Summit on Sep 15, 2026 with Gwynne Shotwell — 'imagine what action one might take' — an expression of intent that rules_secondary explicitly says does NOT qualify. Tesla's March 2026 $2B stake in SpaceX (<1%) is not a controlling-interest transfer.
Calendar: 39 days. A definitive agreement between a public company and a ~$1T private company needs fairness opinions, special-committee process and an 8-K; nothing of that shape is on any public calendar. Polymarket 'announced by Sep 30' sat ~5% (Yahoo Finance, Sep 16).
Already happened?: No.
Precedents consulted: No rows for KXCOMPANYACTIONMERGER. Nearest archetype rows (KXIPOANTHROPIC JUN01/JUL01, NO): Kalshi's corporate-event resolvers read triggers mechanically — filed/intends/hinted never pays; only the defined instrument does.
Resolver profile & risk: Single trigger, tightly defined; Musk-statement carve-out is a dated clarification that protects the NO. Deduction ≈ 2 pts for 'official statements to media subsequently confirmed by the company' ambiguity.
Cleanest tail: Cleanest loss: a surprise definitive-agreement 8-K before Nov 1 — new discrete event. Acceptable shape.
Book / 48h move: Range-bound 7–12¢ for two weeks; 36k contracts on Sep 16 (post-All-In) did not move the close. No 48h dislocation. OI 90.8k, specialist-dense. Entry context: YES bid 10 / ask 11, depth at touch 276 contracts, vol24h 2,130, OI 90,782.

scheduled-process NO side below band (<60¢)

Bank of Canada hikes 25 bp at Oct 28, 2026 meeting?

KXCBDECISIONCANADA-26OCT-H25 · closes 2026-10-28

REJECT — below band. NO costs 56¢ (YES 44/47), under the 60¢ floor; also a live macro-threshold coin flip where our probability is inside the spread — no edge to price

YES condition: For YES to pay, the Bank of Canada must raise the overnight target by exactly 25 bp at its fixed-date Oct 28, 2026 announcement.
Thesis shape: Scheduled-process (a decision at a dated meeting) — Anchor-shaped in form, but the content is a macro call.
Probability (YES): Prob: 45% pre-market → 44% final. Market YES 44/47; NO effective entry 56¢.
Vehicle: The vehicle exists and is dated: Oct 28, 2026 fixed announcement date (BoC schedule). Catalyst chain before it: Sep jobs (~Oct 9), Business Outlook Survey (~Oct 19), Sep CPI (~Oct 20).
Calendar: Sep 2 hold at 2.25% (seventh straight; bankofcanada.ca press release) with Macklem: 'if it takes more than one increase, we're prepared to do that.' Sep 21 Halifax speech (Bloomberg/CMT): flags the risk of 'delaying a rate hike too long'. Market OIS ~50% (Globe and Mail, Yahoo Finance Canada); sell-side base case still hold (Capital Economics' Brown: 'close call').
Already happened?: No.
Precedents consulted: KXCBDECISIONCANADA JUN/JUL/SEP ladders all resolved hold; the Sep-22 note in the store says the NO-hike leg 'only becomes an Anchor-shaped trade when it is inside the 60–84¢ band with the last pre-meeting CPI already printed; at parity with OIS before that print there is no edge.' KXFEDDECISION-26JUL-H25 (NO): a hawkish repricing into a live meeting was a fade — but that repricing peaked at 36–38%, not 50%.
Resolver profile & risk: Mechanical: BoC press release at 09:45 ET. Zero definitional risk (0 pts).
Cleanest tail: Cleanest loss for NO: the BoC hikes — which is the current trend (oil, 3% CPI, Governor pre-signalling) continuing. That is a continuation tail. Would fail Stage 3 even if it were in band.
Book / 48h move: H25 ran 15¢ (Sep 10) → 48¢ last (Sep 23) in two weeks on steady volume; H0 mirror 85 → 57. Book thick on both sides (1.4k NO at 51, 3k YES bid at 55 on the H0 leg). Entry context: YES bid 44 / ask 47, depth at touch 1386 contracts, vol24h 3,402, OI 28,263.

scheduled-process YES side below band (<60¢)

Bank of Canada holds at Oct 28, 2026 meeting?

KXCBDECISIONCANADA-26OCT-H0 · closes 2026-10-28

REJECT — below band. hold leg costs 57¢, under the 60¢ floor; also a yes-side thesis whose 'process-in-motion' step is a meeting whose outcome is the question, and the loss tail (hike) is the trend continuing

YES condition: For YES to pay, the BoC must leave the overnight target unchanged at 2.25% on Oct 28, 2026.
Thesis shape: Scheduled-process; YES side. Fails the side rule (the calendar step exists, but the precedent condition 'resolver pays on that step' is about the decision content, not the step).
Probability (YES): Prob: 55% pre-market → 56% final. Market YES 56/57; YES effective entry 57¢.
Vehicle: Same as H25 leg.
Calendar: Same as H25 leg.
Already happened?: No.
Precedents consulted: KXCBDECISIONCANADA-26SEP-H0: 'Kalshi settles the 0bps leg YES on an unchanged target rate; seven straight holds.' KXFEDDECISION-26JUL-H0 (YES): 'buy the hold rung when a hawkish repricing has knocked it well below par' — but that was 62–64¢ vs 36% hike odds; here the hold is 57¢ vs ~50% OIS. No room.
Resolver profile & risk: Mechanical (0 pts).
Cleanest tail: Loss = hike = trend continues. Continuation tail.
Book / 48h move: 57¢ last; 3k YES bid at 55, 129 at 56. Liquid. Entry context: YES bid 56 / ask 57, depth at touch 129 contracts, vol24h 2,316, OI 28,819.

corporate-event NO side deep-favorite (≥90¢)

Anthropic officially announces IPO before Nov 1, 2026?

KXIPOANTHROPIC-DATE-26NOV01 · closes 2026-11-01

REJECT — deep-favorite band. NO is 94¢ marketable (YES bid 6). WSJ Sep 18 delay-to-November report already collapsed YES from 53¢ to 8¢; nothing left to fade

YES condition: For YES to pay, the SEC must declare Anthropic's S-1 effective, OR the IPO must price, OR an exchange must assign a ticker, before Nov 1, 2026.
Thesis shape: Anchor-shaped (process reaching a defined step). Mechanism if traded: calendar-impossibility / status-quo.
Probability (YES): Prob: 5% pre-market → 4% final. Market YES 6/8; NO effective entry 94¢.
Vehicle: Process is in motion (confidential S-1 filed June; Nasdaq venue picked Sep 13 per Bloomberg) but the dated step moved: WSJ (Sep 18, via Investing.com/PYMNTS) reports the IPO staging pushed to November to show Q3 numbers; Reuters (Sep 5, CNBC) had marketing at mid-October at the earliest. Bloomberg Sep 22: 'derailed IPO plans rattle US market ahead of Anthropic debut'.
Calendar: An effective S-1 or pricing before Nov 1 now requires a public S-1 flip and a roadshow launched within ~3 weeks, against the company's own reported November plan.
Already happened?: No — none of the three triggers has occurred.
Precedents consulted: KXIPOANTHROPIC-DATE-26JUN01/JUL01 (NO): triggers are mechanical; 'filed' ≠ 'confirmed'. KXIPO-26-SPACEX row: a proposed symbol in an S-1 is not 'an exchange has assigned a ticker'; venue-selection headlines are non-events.
Resolver profile & risk: Three explicit triggers; ticker-assignment clause is the only soft edge (–3 pts).
Cleanest tail: Loss = accelerated public S-1 + pricing in late October — a new discrete event, acceptable shape.
Book / 48h move: 53¢ Sep 17 → 8¢ Sep 18 on 37k contracts; 5–8¢ since. The move already happened; NO at 94 is deep-favorite. Entry context: YES bid 6 / ask 8, depth at touch 240 contracts, vol24h 4,687, OI 94,420.

departure/ouster NO side deep-favorite (≥90¢)

Max Miller out as U.S. Representative before Nov 3, 2026?

KXMILLEROUT-26JUL-26NOV03 · closes 2026-11-03

REJECT — deep-favorite band. NO is 96¢ marketable (YES bid 4). Miller is campaigning (attack-ad dispute reported Sep 23, Ohio Capital Journal); no vehicle for an exit before Nov 3

YES condition: For YES to pay, Rep. Max Miller must actually vacate his House seat (resign, be expelled, die) before Nov 3, 2026.
Thesis shape: Anchor-shaped. Mechanism if traded: status-quo.
Probability (YES): Prob: 3% pre-market → 3% final. Market YES 4/6; NO effective entry 96¢.
Vehicle: None. Ohio ballot-replacement window closed Aug 8–9 (Washington Post); Miller stayed on as GOP nominee (WaPo Aug 10); Sep 23 coverage is about his race's attack ads, i.e. he is an active candidate in office. Expulsion needs a two-thirds House vote with no resolution pending.
Calendar: 41 days; the House is in and out of session through October. Nothing scheduled.
Already happened?: No.
Precedents consulted: KXLEAVEWALZ-27JAN01-TWAL-26JUL01 (NO): 'quitting a race, resignation demands and filed impeachment articles all leave the incumbent in office — only resignation, completed removal or death resolves Yes.'
Resolver profile & risk: Standard departure rules_secondary; death → last fair price (–1 pt).
Cleanest tail: Loss = surprise resignation — new discrete event. Acceptable shape.
Book / 48h move: Flat 4–7¢ for two weeks, thin (v24 797). Deep-favorite. Entry context: YES bid 4 / ask 6, depth at touch 170 contracts, vol24h 797, OI 49,889.

confirmation/enact NO side deep-favorite (≥90¢)

Mamdani signs >0 executive orders Sep 20–26, 2026?

KXMAMDANIEO-26SEP26-T0 · closes 2026-09-27

REJECT — continuous-number market. weekly executive-order count (classifier slip); also closes Sep 27 — effectively inside 7d even though expected_expiration is Dec 25

YES condition: For YES to pay, Mamdani must sign at least one non-emergency executive order between Sep 20 and Sep 26, 2026 (a count, resolved by NYC's EO register).
Thesis shape: Count/level market → continuous-number thesis. Stopped at step 2; no probability formed beyond the book.
Probability (YES): Prob: not formed pre-market → 10% final. Market YES 10/11; NO effective entry 90¢.
Vehicle: n/a
Calendar: n/a
Already happened?: n/a
Precedents consulted: None in store for KXMAMDANIEO.
Resolver profile & risk: n/a
Cleanest tail: n/a
Book / 48h move: 7¢ → 12–15¢ Sep 21 (10.6k vol) → 11¢. Not researched further. Entry context: YES bid 10 / ask 11, depth at touch 96 contracts, vol24h 780, OI 7,673.

5. Reject log

Stage-2 rejects carry the side, price and probability we would have used, so the passed-on scoreboard can score them. Stage-0 rejects carry side/price only.

Ticker	Archetype	Side	Entry	P(YES)	Cluster	Reason
KXKASHOUT-26JUL-26NOV01	departure/ouster	NO	90¢	7%	kash-patel-fbi-exit	deep-favorite band
KXCOMPANYACTIONMERGER-27-26NOV01	corporate-event	NO	90¢	7%	tesla-spacex-merger	deep-favorite band
KXCBDECISIONCANADA-26OCT-H25	scheduled-process	NO	56¢	44%	boc-oct-2026-decision	below band
KXCBDECISIONCANADA-26OCT-H0	scheduled-process	YES	57¢	56%	boc-oct-2026-decision	below band
KXIPOANTHROPIC-DATE-26NOV01	corporate-event	NO	94¢	4%	anthropic-ipo-timing	deep-favorite band
KXMILLEROUT-26JUL-26NOV03	departure/ouster	NO	96¢	3%	max-miller-house-exit	deep-favorite band
KXMAMDANIEO-26SEP26-T0	confirmation/enact	NO	90¢	10%	mamdani-eo-count	continuous-number market
KXLOCALLAW-CAAANG-26OCT01	confirmation/enact	NO	100¢	—	anaheim-angels-bill	YES ask 1¢ ≤ 5¢ (Stage-0); NO side would be 100¢ = deep-favorite band
KXAGICO-COMP-26Q3	multi-news-other	NO	99¢	—	agi-announcement	YES ask 2¢ ≤ 5¢ (Stage-0); NO side would be 99¢ = deep-favorite band
KXRETIREMM-26	departure/ouster	NO	96¢	—	mcconnell-exit	YES ask 4¢ ≤ 5¢ (Stage-0); NO side would be 96¢ = deep-favorite band
KXSUEZWEEKLY-26SEP27-T290	legal-action	NO	45¢	—	suez-weekly-transits	continuous-number market: cumulative weekly Suez transit count; also inside 7d (expires Sep 29)
KXSUEZWEEKLY-26SEP27-T260	legal-action	NO	11¢	—	suez-weekly-transits	continuous-number market: cumulative weekly Suez transit count; also inside 7d (expires Sep 29)
KXSUEZWEEKLY-26SEP27-T280	legal-action	NO	34¢	—	suez-weekly-transits	continuous-number market: cumulative weekly Suez transit count; also inside 7d (expires Sep 29)
KXSUEZWEEKLY-26SEP27-T270	legal-action	NO	25¢	—	suez-weekly-transits	continuous-number market: cumulative weekly Suez transit count; also inside 7d (expires Sep 29)
KXHEGSETHOUT-26AUG-26NOV01	multi-news-other	NO	94¢	—	hegseth-dod-exit	close after expiration: close_time Nov 1 is >1h after expected_expiration_time Oct 1
KXHEGSETHOUT-26AUG-26OCT01	multi-news-other	NO	100¢	—	hegseth-dod-exit	YES ask 1¢ ≤ 5¢ (Stage-0); NO side would be 100¢ = deep-favorite band
KXCBDECISIONMEXICO-26SEP24-HOLD	scheduled-process	YES	100¢	—	banxico-sep-2026	inside 7d (expires Sep 24); YES bid 96 ≥ 95
KXGALLEGOOUT-26OCT01	departure/ouster	NO	100¢	—	gallego-exit	YES ask 2¢ ≤ 5¢ (Stage-0); NO side would be 100¢ = deep-favorite band
KXUSAIRANAGREEMENT-27-26NOV	scheduled-process	NO	97¢	—	us-iran-deal	YES ask 4¢ ≤ 5¢ (Stage-0); NO side would be 97¢ = deep-favorite band
KXKASHOUT-26JUL-26OCT01	departure/ouster	NO	100¢	—	kash-patel-fbi-exit	YES ask 1¢ ≤ 5¢ (Stage-0); NO side would be 100¢ = deep-favorite band
KXRETIREPELOSI-26	departure/ouster	NO	97¢	—	pelosi-exit	YES ask 3¢ ≤ 5¢ (Stage-0); NO side would be 97¢ = deep-favorite band
KXIPOANTHROPIC-DATE-26OCT17	corporate-event	NO	99¢	—	anthropic-ipo-timing	YES ask 2¢ ≤ 5¢ (Stage-0); NO side would be 99¢ = deep-favorite band

6. Watchlist — in-band NO shapes just outside the 45-day horizon

Not candidates this run (close Jan 1, 2027 or later). Logged for the next run whose horizon reaches them; prices are today's live quotes.

Ticker	Question	YES bid/ask	NO effective	Vol24h	Note
KXTRUMPADMINLEAVE-26DEC31-PHEG	Hegseth leaves SecDef before 2027	26/27	74¢ (core band)	5,462	Precedent: active war = structural NO for the defence chief. Closes 2027-01-01.
KXTRUMPADMINLEAVE-26DEC31-KPAT	Patel leaves FBI before 2027	16/17	84¢ (core band)	5,800	Same status-quo shape as the Nov 1 rung but in band; 100 days out.
KXTRUMPADMINLEAVE-26DEC31-SWIL	Wiles leaves Chief of Staff before 2027	17/18	83¢	2,147	No vehicle found; not researched this run.
KXTRUMPADMINLEAVE-26DEC31-HLUT	Lutnick leaves Commerce before 2027	16/18	84¢	596	Precedent: absent a named successor, price near base rate.
KXMJSCHEDULE-27	Marijuana rescheduled before 2027	23/24	77¢	6,165	Process-in-motion check needed (DEA/OLC step). Not researched.
KXFEDERALCHARGE-27JAN01-LJAM	Letitia James charged before 2027	11/15	89¢ (thin band, wide spread)	1,283	Spread 4¢; legal-action archetype.

7. Risk profile & execution notes

Capital at risk: $0 of $2,500. Reserve 100% (≥10% required).
What would change the verdict inside the horizon: (a) Patel Nov 1 YES lifting to ≥ 12¢ on a thick book (NO ≤ 88 marketable) with still no primary-source vehicle → thin-favorite NO at half size (5% = $125, ≤ 1/3 visible depth); (b) BoC Oct H25 falling back to ≤ 40¢ after a soft Sep CPI on ~Oct 20, per the store's Sep-22 note — but its loss tail stays a continuation tail, so Anchor would still pass; (c) Tesla/SpaceX YES ≥ 12¢ on a Musk hint with no 8-K → same thin-favorite logic as Patel.
Do not: rest NO orders at 88–89 on Patel or Tesla/SpaceX to manufacture a band fill — that is 'inside spread' behind 170–384 contracts of existing bids, and a fill there most likely means the news went the wrong way.
Existing pick untouched: KXCRYPTOSTRUCTURE-26JAN-OCT NO (not in this allowlist; closes Oct 1, inside 7d in any case).

8. Sources consulted this run

Kalshi public API: /markets, /markets/{ticker}, /markets/{ticker}/orderbook, /series/{series}/markets/{ticker}/candlesticks (60-min, 14 days). DB query API: resolution_precedents (162 rows; 12 Bank of Canada rows, 3 KXKASHOUT-adjacent rows), signals (prior Anchor runs Sep 8, 21, 22).
Bank of Canada press releases (bankofcanada.ca): Sep 2, 2026 hold at 2.25%; Sep 2 press-conference opening statement; Sep 21 Halifax speech via Bloomberg / Canadian Mortgage Trends; Globe and Mail and Yahoo Finance Canada on ~50% OIS hike odds; Capital Economics (Stephen Brown) hold base case.
Patel: CBS News / Washington Post / NBC / CNN coverage of the Sep 15, 2026 Senate Judiciary hearing; Al Jazeera Sep 15; New Republic and MS NOW (late-2025 'Trump weighing replacement' reports, no follow-through).
Anthropic: Reuters via CNBC (Sep 5, mid-October marketing); WSJ via Investing.com / PYMNTS (Sep 18–20, staging pushed to November); Bloomberg Sep 22.
Tesla/SpaceX: Benzinga / Teslarati / Yahoo Finance on the Sep 15 All-In Summit remarks; Yahoo Finance on Polymarket ~5% by Sep 30 (Sep 16).
Miller: Washington Post (Aug 9–10, ballot window closed, remains nominee); Ohio Capital Journal / News 5 Cleveland (Sep 23, attack-ad dispute).