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

Date: 2026-09-29
Source: https://kalshi-audits.pages.dev/2026-09-29-kalshi-politics-mispricing-audit-anchor

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

Report date 2026-09-29 · strategy=Anchor · audit_version=anchor · horizon 45 days · capital $2,500 · focus category: Politics (allowlist-wide scope)
Picks1Logged rejects14Capital deployed$1847.4% · cash 92.6%Stage-0 survivors6of 869 open markets / 115 series

## 1. Summary

The Stage-0 allowlist (120 series, 115 with open markets) yields **869 open markets**, of which only **6** survive the mechanical cuts inside the 45-day horizon (close ≤ Nov 13; vol24 ≥ 500; spread ≤ 5¢; 5¢ < YES < 95¢; ≥ 7 days to expiry). Of those six, one produces an Anchor-shaped NO pick in the core band: **NO on “Anthropic confirms IPO before Nov 1” at 80¢**. Two are deep favorites (NO ≥ 90¢: Patel out by Nov 1, Iran deal by Nov 1), two are the Bank of Canada October ladder (a directional macro call — below band on NO, yes-side thesis on the hold leg), and one (diesel export ban) has no NO edge because my own probability sits above the book. Cash is the position for 92.6% of capital, which is the expected Anchor outcome in an efficient week.

**Disclosure.** The analyst model is built by Anthropic. The pick concerns the timing of Anthropic's IPO. The thesis uses only public sources named below (EDGAR, Reuters/CNBC, WSJ via secondary coverage, Fortune, Polymarket, Kalshi precedents); the model has no non-public information about the offering (training cutoff June 2026 predates every fact used). Readers should weigh that optic themselves.

## 2. Pick

BUY NO @ 80¢mechanism: no-vehicle (+ overreaction-fade)corporate-eventconviction: MEDIUMcluster: anthropic-ipo-timing

### KXIPOANTHROPIC-DATE-26NOV01 — Anthropic confirms IPO (S-1 effective, priced, or ticker assigned) before Nov 1, 2026?

**YES condition:** For YES to pay, the SEC must declare Anthropic's Form S-1 effective, _or_ the IPO must be priced, _or_ a securities exchange must assign it a ticker, before Nov 1 2026 04:59Z (Oct 31 midnight ET), resolved by Kalshi from SEC/exchange records (rules_secondary, verbatim: “An IPO is confirmed if 1) the SEC declares the company's Form S-1 effective OR 2) the IPO is priced OR 3) a securities exchange has assigned a ticker to it”).

**Thesis.** No vehicle: no public S-1 exists on EDGAR as of Sep 29, so the mandatory 15-day pre-roadshow clock has not started; Reuters (Sep 19, restated Sep 28) has the debut slipping past the Nov 3 midterms; today's 6¢→20¢ jump is the prospectus LEAK, which is not a resolution event.

**Probability:** 11% pre-market → **10% final** (YES). NO fair value 90¢ vs 80¢ offered → edge 10¢, EV +12.5% on cost.

**Vehicle / calendar / already-happened checks.**

- **Vehicle:** none. EDGAR full-text search (S-1, S-1/A, DRS; Jun 1–Sep 29) returns only third-party filers that mention Anthropic (SpaceX, Oura); the EDGAR current-filings S-1 feed for Sep 25–29 lists no Anthropic filing. CNBC/Reuters (Sep 29): the paperwork “had still not been publicly disclosed”; Fortune (Sep 29): the prospectus “was leaked to Reuters”.
- **Calendar:** the SEC's draft-registration process requires the public filing at least 15 days before a roadshow. No flip by Sep 29 → earliest roadshow ~Oct 14–15 → a pre-Oct-31 pricing needs a flip within days and a compressed roadshow. Reuters (Sep 19, restated in its Sep 28 prospectus story): the debut “is likely to be pushed to after the November US midterm elections”; WSJ (Sep 18, via secondary coverage): shift from October to November to incorporate Q3 results (Q3 closes Sep 30). Timeline has slipped three times (Aug → mid-Oct marketing → “days before midterms” → after midterms).
- **Already happened?** No. Trigger 1/2/3 all unmet; a confidential draft and a leaked draft are neither.

**Market context (opened after the pre-market number).** NOV01 traded 53–56¢ Sep 15–17, collapsed to 5–8¢ on Sep 18 (37k contracts) on the post-midterms reporting, sat 4–7¢ Sep 20–28, then jumped 6¢ → 35¢ intraday today (Sep 29, 11.3k contracts) on the prospectus-leak headlines, settling 20/21. Polymarket at the same time: “Anthropic IPO by Oct 31” **4.5%**, “IPO in October” **3.9%**, “by Nov 15” 35%, “in November” 61.5% (24h volume $24k / $2k / $12k / $7k). Kalshi's DEC01 rung is 56/70 and JAN01 76/77, consistent with a November pricing. The precedent store documents this exact pattern twice (Jun 2 confidential-filing headline; Sep 13 Nasdaq-venue headline): “filing-adjacent headline residue” that decays because it changes nothing in the rulebook triggers.

**Resolver profile.** Resolver: Kalshi, mechanical three-trigger definition. Precedents: KXIPOANTHROPIC JUN01 and JUL01 both settled NO after the confidential filing (“'filed' and 'confirmed' are separated by months of process”); KXIPO SpaceX settled on the Jun 11 pricing, not the May 20 public S-1 with proposed symbol (“a proposed symbol in a public S-1 is not 'a securities exchange has assigned a ticker'”). Risk priced: −3 pts for an aggressive “ticker assigned” reading if Nasdaq formally approves the listing before pricing. No clarification in rules_secondary cuts against NO.

**Cleanest tail (new discrete event required):** Anthropic publicly files on EDGAR this week and launches an accelerated roadshow to price Oct 27–30, ahead of the midterms — a reversal of the reported slippage, not its continuation. Priced at ~10%. Continuation of the current trend (further slippage) favors NO.

**Precedents consulted:** KXIPOANTHROPIC-DATE-26JUN01 (no), KXIPOANTHROPIC-DATE-26JUL01 (no), KXIPOANTHROPIC-DATE-26SEP01 (edge note re: Kalshi/Polymarket parity, “only a dated S-1 public flip or a roadshow launch changes it”), KXIPO-26-SPACEX (settled on pricing, Jun 15).

**Sources:** EDGAR FTS (efts.sec.gov, S-1/S-1A/DRS Jun 1–Sep 29) · EDGAR current S-1 feed (Sep 25–29) · Reuters via Yahoo Finance, Sep 29 (“seen by Reuters… likely to be pushed to after the November US midterm elections, Reuters previously reported”) · CNBC Sep 29 (Reuters) · Fortune Sep 29 (“leaked to Reuters”) · GuruFocus Sep 21 summarizing Reuters Sep 19 · Motley Fool Sep 26 (October → November) · CNBC Sep 5 (Reuters: public S-1 late Sep, marketing mid-Oct earliest) · Polymarket gamma API, Sep 29 16:45Z · Kalshi public API market/orderbook/candlesticks, Sep 29 16:40Z.

## 3. Position sizes & limits

| Ticker | Side | Limit | Contracts | Cost (= max loss) | % of capital | Depth at/below limit | Depth cap (1/3) | Conviction cap |
|---|---|---|---|---|---|---|---|---|
| KXIPOANTHROPIC-DATE-26NOV01 | NO | 80¢ | 230 | $184 | 7.4% | 690 (YES bids at 20¢) | 230 | $250 (10%, MEDIUM) |

Flat sizing: 10% of capital max per pick regardless of conviction; the depth cap binds (690 resting YES bids at 20¢ → 230 contracts). Slippage 0¢ at an 80¢ limit; a 81¢ limit would add 1,000 contracts of depth but is not needed. Reserve held: $2,316 (92.6%).

## 4. EV for the record

| Ticker | Side | Price | Model P(YES) | Model P(side) | Edge | EV/contract | EV % of cost | Max win | Max loss |
|---|---|---|---|---|---|---|---|---|---|
| KXIPOANTHROPIC-DATE-26NOV01 | NO | 80¢ | 10% | 90% | +10¢ | +10.0¢ | +12.5% | $46 | $184 |

## 5. Cluster table

| Cluster | Deployed leg | Runner-ups / same-cluster rejects |
|---|---|---|
| anthropic-ipo-timing | KXIPOANTHROPIC-DATE-26NOV01 NO | OCT24 (spread >5¢, deep-favorite), OCT17, OCT10 (ask ≤5¢) |
| diesel-export-ban | — | KXDIESELEXPORTBAN-26 (inside spread) |
| patel-fbi-exit | — | NOV01 (deep-favorite band), OCT01 (inside 7d) |
| iran-nuclear-deal | — | NOV (deep-favorite band), OCT (inside 7d) |
| boc-oct-decision | — | H25 NO (below band), H0 YES (yes-side thesis) |
| hegseth-exit / zelenskyy-exit / mcconnell-resign / agi-announce | — | Stage-0 cuts (inside 7d or ask ≤5¢) |

## 6. Risk profile

- **Single-name, single-cluster exposure:** one position, $184 at risk (7.4% of capital). Portfolio max drawdown = that cost.
- **Way to lose:** a public EDGAR filing in the next ~7 days followed by a roadshow launch around Oct 14 and a pricing in the last week of October. Watch EDGAR (form S-1 by “Anthropic, PBC”) and Kalshi's OCT24/NOV01 rungs; a public flip before ~Oct 7 is the exit signal — sell NO or hold depending on the reported pricing date.
- **Resolver risk:** low; priced at 3 pts (ticker-assignment reading). Rules_secondary is unchanged since listing.
- **Liquidity:** vol24 11.6k (specialist-dense; edge demanded is specific: EDGAR absence + 15-day rule + Polymarket 4.5% + Reuters post-midterm stance). Exit liquidity exists at the NO bid stack (0.70–0.79, 3.4k contracts).
- **Hold horizon:** 33 days to close; Kalshi resolves early on any trigger.

## 7. Execution notes

- Place BUY NO 230 @ 80¢ (rests against/ fills the 690 YES bids at 20¢). Do not chase above 84¢ (edge < 6 pts).
- If the price has reverted toward 90¢+ NO (YES ≤ 10¢) before fill — as it did Sep 20–28 — the deep-favorite band rule applies: stand down.
- Re-check EDGAR and the Kalshi rules_secondary field before each session; a dated public S-1 is the only fact that changes this thesis.

## 8. Rejected (logged for the passed-on scoreboard)

| Ticker | Side | Price (side) | P(YES) | Cluster | Reason |
|---|---|---|---|---|---|
| KXDIESELEXPORTBAN-26 | NO | 83¢ | 25% | diesel-export-ban | inside spread: model 25% YES (Politico Sep 23: 90-day ban plan being prepared; Trump Sep 27 'very seriously… we might do it'; diesel at a record $6.50) sits ABOVE the 17–21¢ book, so NO at 83¢ has no edge; the rulebook never defines 'ban' vs partial restrictions (Wright: 'blanket ban not being considered'), a live resolver risk; book is headline-driven (78¢ intraday Sep 23). |
| KXKASHOUT-26JUL-26NOV01 | NO | 93¢ | 5% | patel-fbi-exit | deep-favorite band: NO costs 91–93¢ (YES 7/9); status-quo thesis is sound (deputy Bailey left Sep 28, contentious Senate hearing, but no vacancy signal — KXKASHOUT precedents: nothing short of an actual vacancy pays) yet max win ≤9% and the 90¢+ cohort showed no edge. |
| KXUSAIRANAGREEMENT-27-26NOV | NO | 96¢ | 4% | iran-nuclear-deal | deep-favorite band: NO costs 94–96¢; The National (Sep 29) has only quiet indirect talks in New York with 'breakthrough elusive' and no draft text, and the 06/17/26 clarification requires a written instrument with concrete nuclear limits AND sanctions relief — but ≤6% max win is outside the Anchor band. |
| KXCBDECISIONCANADA-26OCT-H25 | NO | 59¢ | 45% | boc-oct-decision | below band: NO costs 58–59¢ (<60¢) and the thesis is a directional macro call (markets 'coin flip, narrowly favour a hike' per Globe and Mail Sep 18; RBC/Desjardins/Capital Economics expect a hold, National Bank/Scotiabank a hike) — no coin-flip lane in Anchor. |
| KXCBDECISIONCANADA-26OCT-H0 | YES | 62¢ | 54% | boc-oct-decision | yes-side thesis: the Oct 28 meeting is on the published calendar, but the question is the DIRECTION of the decision (a macro threshold call), not whether a process step completes — neither already-happened nor process-in-motion applies; NO on this leg would cost 38–39¢ (below band). |
| KXIPOANTHROPIC-DATE-26OCT24 | NO | 97¢ | 3% | anthropic-ipo-timing | cluster runner-up: same no-vehicle thesis as the NOV01 leg but 3/25¢ book (spread >5¢, Stage-0 cut) and NO ≥90¢ (deep-favorite band). |
| KXIPOANTHROPIC-DATE-26OCT17 | NO | 98¢ | 1% | anthropic-ipo-timing | deep-favorite band / YES ask ≤5¢ (Stage-0 cut); calendar-impossible given the 15-day flip rule, but nothing left to earn. |
| KXIPOANTHROPIC-DATE-26OCT10 | NO | 99¢ | 1% | anthropic-ipo-timing | deep-favorite band / YES ask ≤5¢ (Stage-0 cut). |
| KXUSAIRANAGREEMENT-27-26OCT | NO | 100¢ | 1% | iran-nuclear-deal | inside 7d and YES ask ≤1¢ (Stage-0 cut). |
| KXKASHOUT-26JUL-26OCT01 | NO | 100¢ | 1% | patel-fbi-exit | inside 7d and YES ask ≤2¢ (Stage-0 cut). |
| KXHEGSETHOUT-26AUG-26OCT01 | NO | 100¢ | 1% | hegseth-exit | inside 7d (expires in 1.9 days) and YES ask ≤1¢ (Stage-0 cut). |
| KXZELENSKYYOUT-26OCT01 | NO | 100¢ | 1% | zelenskyy-exit | inside 7d (expires in 1.9 days) and YES ask ≤3¢ (Stage-0 cut). |
| KXRETIREMM-26 | NO | 97¢ | 2% | mcconnell-resign | deep-favorite band / YES ask ≤5¢ (3.0/3.3¢; Stage-0 cut) despite 17.8k vol24. |
| KXAGICO-COMP-26Q3 | NO | 99¢ | 1% | agi-announce | inside 7d and YES ask ≤2¢ (Stage-0 cut). |

Existing live pick KXCRYPTOSTRUCTURE-26JAN-OCT (NO) is outside the allowlisted series and was not re-researched. Stage-0 cut 863 markets: 830 close beyond the 45-day horizon or trade < 500 contracts/24h; the remaining 33 within-horizon cuts were YES ask ≤ 5¢, spread > 5¢ or inside-7d and the material ones are logged above.

## 9. Candidates that reached Stage 2 — diligence notes

### KXDIESELEXPORTBAN-26 — NO 83¢, P(YES) 25% → reject (inside spread)

YES condition: a US ban on diesel exports must have _gone into effect_ after issuance and before Nov 3, as reported by a listed Source Agency. Vehicle: Politico (Sep 23) — administration preparing a 90-day export-ban plan; Axios (Sep 22) — Trump backs a ban; Trump (Sep 27, Fox): “We're seriously thinking about it… We might do it.” Counter: Energy Secretary Wright (Sep 23–24) — a blanket ban is not being considered, voluntary adjustments preferred; API, Chamber/Business Roundtable letter (Sep 24) opposed. Calendar: EPCA short-supply authority can take effect within days, so nothing rules YES out. Price path: 21¢ → 78¢ intraday Sep 23 (36.6k) → 12¢ Sep 26 → 22¢ Sep 28 → 17/21 now. My 25% is above the book; a NO at 83¢ has negative expectancy. Resolver risk: “ban” undefined against a likely partial/licensed restriction. Precedents for this series: none (KXTARIFFCHECKS lessons on 'reported-fact' gates noted).

### KXKASHOUT-26JUL-26NOV01 — NO 93¢, P(YES) 5% → reject (deep-favorite band)

YES condition: Patel must vacate the FBI Director role before Nov 1 (actual departure date, not announced intent). News: deputy director Andrew Bailey out Sep 28 (Axios); contentious Senate oversight hearing; NBC on reduced briefings. No vacancy vehicle. Precedents (KXKASHOUT JUN01/JUL01, both NO): far rungs systematically overpriced; only an actual vacancy pays. Right thesis, wrong band.

### KXUSAIRANAGREEMENT-27-26NOV — NO 96¢, P(YES) 4% → reject (deep-favorite band)

YES condition: a formal written US–Iran agreement with concrete, verifiable nuclear limits AND sanctions relief, agreed/signed before Nov 1. The National (Sep 29): quiet indirect talks in New York, “breakthrough elusive,” dispute over who takes delivery of the 440 kg of 60% uranium, no draft text, no scheduled round. Precedents (JUN, JUL rungs, both NO; 06/17/26 clarification): an MoU deferring nuclear substance does not pay. 33 days is short for a text; but ≤6% max win is outside the band.

### KXCBDECISIONCANADA-26OCT (H25 / H0) — reject (below band / yes-side thesis)

YES condition (H25): the Bank of Canada must raise the overnight target 25bp at the Oct 28 fixed announcement date. Rate 2.25% since late 2025; markets “a coin flip, narrowly in favour of a hike” (Globe and Mail, Sep 18); RBC, Desjardins, Capital Economics expect a hold with a Q1-2027 hike; National Bank and Scotiabank forecast an October hike; the October MPR will raise inflation forecasts on oil. My 45% hike / 54% hold is inside the book. Precedents (12 rows, July ladder): outer rungs uninformative; the July hold was fully priced. Anchor has no coin-flip lane and no directional-macro YES lane.

## 10. Method & data notes

Universe from /tmp/out/targets.json (generated 2026-09-29T16:37Z, min OI 5,000). Live data from the Kalshi public API (markets, orderbook, candlesticks period 60) at 16:40Z; all prices read from *_dollars / *_fp fields. Precedents from resolution_precedents (162 rows; 3 for KXIPOANTHROPIC, 2 KXUSAIRANAGREEMENT, 2 KXKASHOUT, 12 KXCBDECISIONCANADA, 0 KXDIESELEXPORTBAN). Probabilities were committed before the orderbook and candlesticks were opened; the Anthropic number moved 11% → 10% after the Polymarket cross-check. Anchor rules applied in order: side rule → continuous-number → continuation tail → rules-reading edge → band → one leg per cluster → flat sizing with depth cap.