# Kalshi Politics — Mispricing Audit (v4, archetype-targeted) — 2026-08-19

Date: 2026-08-19
Source: https://kalshi-audits.pages.dev/2026-08-19-kalshi-politics-mispricing-audit-v4

---
# Kalshi Politics — Mispricing Audit

v4 methodology (archetype-targeted, precedent-fed) · Report date 2026-08-19 · Horizon 45 days · Capital $2,500

## Summary

Universe (targets file)116 series
Open mkts in horizon96
Stage-0 survivors5
Picks1
Logged rejects5
Deployed$174.75
Cash93%

A thin week, and the discipline says that is fine. Of 96 open markets in the pre-classified
target series closing by Oct 3, only five cleared the hard liquidity/spread/price floors (98 of 136 historical
picks entered zero-volume books, and that cohort lost money — the floor exists for a reason). Of the five, two
died mechanically on the entry bands, two died on honest two-sided uncertainty, and one — **NO on Anthropic
IPO-confirmed-before-Oct-1 at 75¢** — carries a specific, precedent-backed, rules-text edge worth deploying at
MEDIUM size. 93% of capital stays in cash.

## Pick 1 of 1 — BUY NO · Anthropic IPO confirmed before Oct 1

BUY NO @ 75¢
MEDIUM conviction
corporate-event · Tier 2
core band 60–84¢

### KXIPOANTHROPIC-DATE-26OCT01 · closes 2026-10-01

**Market:** "When will Anthropic officially announce an IPO?" — before Oct 1, 2026 leg.
  YES 25/26¢, vol24h 5,344 (specialist-dense), OI 115,124.

**Resolver profile.** Kalshi resolves this series on three purely mechanical triggers, any
  one of which fires YES immediately: (1) SEC declares the S-1 _effective_, (2) the IPO is _priced_,
  (3) a securities exchange _assigns a ticker_. The headline word "announce" in the title is a trap — the
  rules text, and this resolver's own record, say announcements do not count.

**Evidence (primary sources).** Anthropic confidentially submitted a draft S-1 on
  **June 1, 2026** (Anthropic press release). As of **Aug 19**, SEC EDGAR company search returns
  **no public Anthropic registrant and no public S-1** — verified directly this run. All reporting
  (Bloomberg, CNBC, seekingalpha, Aug 18) points to an **October 2026 Nasdaq listing** led by Goldman/JPM/MS.
  On a normal timeline — public flip ~3–4 weeks before listing, roadshow, pricing the night before trading —
  effectiveness and pricing land in October, after this market's cutoff.

**Probability.** Pre-market **NO 88%** (YES 12%) → final **NO 84%**, blended toward
  a liquid specialist book sitting at 25¢ YES. Resolver-risk deduction: −4 pts for the "exchange has assigned a
  ticker" clause — if the public flip lands in September naming a proposed Nasdaq symbol, a loose reading could
  fire early. The strict reading (an application "under the symbol X" is not exchange assignment) and the SpaceX
  precedent (all three triggers fired together at pricing, not at the flip) both cut against that, hence a small
  deduction rather than a reject. World risk — pricing pulled forward to Sep 29–30 in the "race against OpenAI"
  — is why this is MEDIUM, not HIGH.

**48h price-move check.** YES moved 20¢ → 25¢ on Aug 18–19. Explained: a headline burst
  (super-voting shares for CEO Amodei, pre-IPO credit facility past $10B, "Anthropic steps up IPO prep") — all
  IPO-_preparation_ news, none of it a resolution trigger. This is the identical announcement-is-not-resolution
  dynamic the June precedent documented, when 130k contracts repriced YES 5¢→1–2¢ after the confidential-filing
  headline failed to resolve anything. The move is the mispricing, not a warning.

Precedents consulted:
    KXIPOANTHROPIC-DATE-26JUN01 resolved NO — "a confidential draft S-1 lacking price and share count leaves
    all three triggers unmet";
    KXIPOANTHROPIC-DATE-26JUL01 resolved NO — "'filed to go public' is not a resolution event; treat 'filed'
    and 'confirmed' as separated by months of process";
    KXIPO-26-SPACEX resolved YES — "the rulebook fires on the earliest of three mechanical triggers…
    all three were met Jun 11–12", i.e. at pricing time, not at the public flip.

| Entry context | Value | Sizing | Value |
|---|---|---|---|
| YES bid / ask | 25¢ / 26¢ | Conviction cap (MEDIUM, 7%) | $175.00 |
| NO limit (crosses YES bid) | 75¢ | Depth cap (⅓ × 1,005) | 335 contracts |
| Depth at/below limit | 1,005 contracts | **Recommended contracts** | **233** |
| Volume 24h / OI | 5,344 / 115,124 | Cost = max loss | $174.75 |
| Slippage vs last | ≤1¢ ✓ | Max win | $58.25 |

## Logged rejects

NO @ 52¢ consideredconfirmation/enact · Tier 1

### KXGOVTFULLFUND-26AUG-26SEP26 — funding law before Sep 26

**Prob: 42% YES pre-market → 46% final.** The most interesting market of the week, and an honest pass.
  Both chambers passed CRs unusually early — the House's H.R. 9770 (through Dec 4) on July 21, 220–205; the Senate's
  version (through Dec 11, plus a rider delaying the OMB grants rule) on Aug 8, **90–6** — so only House concurrence
  and a signature remain, and midterm pressure favors clearing the deck early. Against that: the White House opposes
  the Senate's OMB rider, the House GOP bill omits it, Johnson faces conference pressure to ping-pong, and in the last
  decade only FY2015 (Sep 19, 2014) beat a Sep 26 cutoff — Congress spends every day the deadline gives it whenever
  anything is contested. **Reject: 40–59¢ band admits HIGH conviction only; this is MEDIUM both ways.**

YES @ 88¢ consideredconfirmation/enact · Tier 1

### KXGOVTFULLFUND-26AUG-26OCT01 — funding law before Oct 1

**Prob: ~91% YES.** Probably the right side, but the market failed the Stage-0 floors
  (vol24h 331 < 500; spread 6¢ > 5¢) and is the cluster runner-up to the Sep 26 leg in any case.
  **Reject: liquidity floor + one leg per cluster (fy27-cr-timing), and no leg deploys.**

YES @ 85¢ consideredscheduled-process · Tier 2

### KXCBDECISIONJAPAN-26SEP17-H25 — BoJ hikes exactly 25bps Sep 17–18

**Prob: 72% YES pre-market → 74% final.** The Aug 14 Reuters sources story ("strong chance" of a
  September hike, possibly at a faster pace) is already the price at 82–85¢. The same story's "more aggressive pace"
  language is a live tail against this exact-25bp leg (a 50bp move resolves it NO), the meeting is a month out, and
  the precedent store's closest analogue — KXFEDDECISION-26JUL-H25, where a hike priced to
  36–38% resolved NO — warns against paying up for leak-driven pricing into a central-bank meeting. YES at 85¢ needs
  ~90%+ true probability at the half-size thin-favorite band; I have 74%. NO at ~18¢ effective is below the 40¢ floor.
  **Reject: no edge on either side.**

NO @ 92¢ considereddeparture/ouster · Tier 1

### KXKASHOUT-26JUL-26OCT01 — Patel out as FBI Director before Oct 1

**Prob: ~8% YES.** NO effective entry is 91–92¢ (≥90¢ is the hard reject band — the record's worst
  loss is the permanent memorial) and YES at 9¢ is below the 40¢ floor. **Reject: deep-favorite band, both sides
  mechanically excluded.**

NO @ 95¢ consideredcorporate-event · Tier 2

### KXCOMPANYACTIONMERGER-27-26OCT01 — Tesla/SpaceX definitive merger agreement before Oct 1

**Prob: ~3% YES.** Rules require an unambiguous, definitive signed agreement (Musk speculation
  explicitly excluded), so YES is a tail — but the trade dies mechanically anyway: NO at 93–95¢ is in the ≥90¢ reject
  band and YES at 7¢ is under the 40¢ floor. **Reject: band.**

## Portfolio tables

### Position sizes & limits

| Ticker | Side | Limit | Contracts | Cost (max loss) | % of capital |
|---|---|---|---|---|---|
| KXIPOANTHROPIC-DATE-26OCT01 | NO | 75¢ | 233 | $174.75 | 7.0% |
| **Cash** | **$2,325.25** | **93.0%** |  |  |  |

### EV for the record

| Ticker | Entry | Model prob | Edge | EV/contract | EV (position) |
|---|---|---|---|---|---|
| KXIPOANTHROPIC-DATE-26OCT01 (NO) | 75¢ | 84% | +9¢ | +9¢ (12%) | +$20.97 |

### Clusters (one leg per cluster)

| Cluster | Deployed leg | Runners-up (logged rejects) |
|---|---|---|
| anthropic-ipo-timing | KXIPOANTHROPIC-DATE-26OCT01 NO | SEP01 leg cut at Stage 0 (ask 1¢) |
| fy27-cr-timing | none deployed | SEP26 leg (band/conviction), OCT01 leg (liquidity floor) |
| boj-sep26 | none deployed | H25 leg (no edge) |

### Conviction exposure & risk profile

| Bucket | Cost at risk | % of capital | Notes |
|---|---|---|---|
| HIGH | $0.00 | 0% | nothing cleared the HIGH bar this week |
| MEDIUM | $174.75 | 7.0% | single NO position, core band, specialist book with a specific rules edge |
| Cash / reserve | $2,325.25 | 93.0% | well above the 10% reserve floor; 40–70% cash is expected in an efficient week — this one is thinner than that, and that is the honest read |

### Execution notes

- Buy NO at a 75¢ limit (crosses the resting 1,005-contract YES bid at 25¢). Do not chase past 76¢ — at 78¢+ the edge net of the ticker-clause risk is gone.
- Watch for the public S-1 flip on EDGAR (likely early-to-mid September). A flip naming a proposed Nasdaq symbol will spike YES on headlines; per the June precedent and the strict rules reading it is _not_ a trigger. It is also the natural exit-review point: if the flip discloses a confirmed September pricing schedule, cut the position.
- The market can close early on any trigger; expiration Oct 8 covers a late resolver read.

## Method note (v4)

Universe restricted to the pre-classified targets file (116 series; Tier 1 worked first). Hard
floors: vol24h ≥ 500 live, spread ≤ 5¢, 5¢ < YES ask, YES bid < 95¢, no existing-pick tickers. Probabilities
committed before opening the book, then blended. Precedent store consulted per candidate (same-series rows existed
for KXIPOANTHROPIC; archetype rows for the BoJ ladder via KXFEDDECISION/KXCBDECISIONCANADA; none for KXGOVTFULLFUND —
cold derivation used). Entry bands: ≥90¢ and <40¢ hard rejects; 40–59¢ HIGH-only; 85–89¢ half size. One leg per
cluster. Sizing EV-agnostic: HIGH 15% / MEDIUM 7% of capital, depth-capped at ⅓ of visible contracts at the limit.

Sources: Anthropic press release (confidential draft S-1, Jun 1, 2026); SEC EDGAR company search
(no public S-1, checked Aug 19, 2026); Bloomberg Aug 18 (super-voting shares; credit facility); CNBC Aug 18;
congress.gov / GovTrack (H.R. 9770 passed House Jul 21, 2026); Senate CR passed 90–6 Aug 8, 2026 (CBS News, Sen. Cramer
press release, NACo); Forbes/The Hill/GovExec (OMB grants-rule rider); Reuters via Investing.com/Yahoo Aug 14 (BoJ
September hike sources story); FXStreet (Reuters July poll, 86% see 1.25% by end-Q4).