Kalshi Politics — Mispricing Audit (Sentinel, managed)
1. How this was researched
Mode: category-match. lower(trading_events.category) = 'politics' matched 307 active markets closing within 45 days on the live Kalshi mirror, so no theme-keyword fallback was needed. Stage-1 mechanical screens (24h volume ≥ 1,500 contracts, bid–ask spread ≤ 5¢, not fully priced, not announcer-noise, not already a live feed pick) cut 307 → 34. Wholesale cuts of noise families (Truth Social post-count ladders, Getty-photo-day ladders, tight RealClearPolitics approval bands) and of six contracts already live in the feed (KXTRUMPATTEND, KXKASHOUT, KXLAGODAYS-0, KXKNESSET, KXFISAEXTEND, KXMAXSHIPSHORMUZ-AL50) left 15 candidates for full Stage-2 diligence: the Strait-of-Hormuz shipping complex (9 markets), the CLARITY Act Senate-vote market, a House budget-resolution market, Trump/UFC 330, Trump World Cup attendance, and two feed-adjacent complements (Mar-a-Lago trips = 1, Kash Patel announces departure).
For each candidate I pulled the resolution rules and 14-day intraday history from the mirror, live top-of-book from the Kalshi public API, and — because the entire Hormuz complex resolves against a single external dataset — the resolver itself: the IMF PortWatch daily chokepoint feed (portid chokepoint6, field n_total), queried directly via its public ArcGIS API. News diligence used primary wire reporting on the July 6–13 US–Iran re-escalation and Congressional-schedule reporting for CLARITY and the budget resolution. Every pick's probability is a Stage-2.5 forecast of the resolver's verdict at resolution time, with resolver risks priced as explicit point deductions.
2. Markets reviewed
| Ticker | Question (short) | YES bid/ask ¢ | Vol 24h | Verdict |
|---|---|---|---|---|
| KXHORMUZWEEKLY-26JUL19-T100 | Hormuz transit calls >100, Jul 13–19 | 28 / 30 | 6,587 | PICK — BUY NO |
| KXHORMUZWEEKLY-26JUL19-T75 | Hormuz transit calls >75, Jul 13–19 | 49 / 56 | 2,430 | REJECT — band screen |
| KXHORMUZWEEKLY-26JUL19-T50 | Hormuz transit calls >50, Jul 13–19 | 70 / 73 | 1,685 | REJECT — tail band, no edge |
| KXHORMUZWEEKLY-26JUL12-T125 | Hormuz transit calls >125, Jul 6–12 | 73 / 75 | 13,510 | REJECT — AIS disadvantage |
| KXHORMUZWEEKLY-26JUL12-T150 | Hormuz transit calls >150, Jul 6–12 | 19 / 24 | 32,385 | REJECT — AIS disadvantage |
| KXHORMUZMAX-26JUL12-JUL07 | Jul 7 the max-transit day of Jul 6–12 | 82 / 85 | 4,023 | REJECT — AIS disadvantage |
| KXHORMUZMAX-26JUL12-JUL06 | Jul 6 the max-transit day of Jul 6–12 | 14 / 16 | 8,470 | REJECT — AIS disadvantage |
| KXMAXSHIPSHORMUZ-26JUL31-AL60 | Max daily Hormuz transits ≥60 in July | 5 / 6 | 4,378 | REJECT — tail, no lottery slot |
| KXHORMUZNORM-26MAR17-B260815 | Hormuz 7-day MA >60 before Aug 15 | 4 / 6 | 16,121 | REJECT — fully priced, feed near-dupe |
| KXCLARITYVOTE-26JUL-AUG08 | Senate recorded vote on CLARITY Act before Aug 8 | 65 / 66 | 1,783 | REJECT — edge eaten (+24¢ in 48h) |
| KXHBUDGETRES-26JUN-26AUG01 | Budget resolution passes House before Aug 1 | 7 / 10 | 4,595 | REJECT — deep favorite, ≤3¢ residual |
| KXTRUMPUFC-26AUG-DJT | Trump attends UFC 330 (Aug 15, Philadelphia) | 12 / 14 | 10,286 | REJECT — tail, speculative |
| KXTRUMPWORLDCUP-26-DJT-1 | Trump attends ≥1 World Cup match | 91 / 92 | 11,162 | REJECT — duplicates live feed pick |
| KXLAGODAYS-26JUL-1 | Trump makes exactly 1 Mar-a-Lago trip in July | 6 / 8 | 26,804 | REJECT — complement of live feed pick |
| KXKASHANNOUNCEOUT-26APR-AUG01 | Kash Patel announces departure before Aug 1 | 6 / 8 | 1,572 | REJECT — correlated with live feed pick |
3. Top picks — detail and thesis
Pick 1 — KXHORMUZWEEKLY-26JUL19-T100 · BUY NO @ 0.73 · HIGH CONVICTION
Question Will there be more than 100 transit calls through the Strait of Hormuz from Jul 13 to Jul 19, 2026?
Current price NO 70/72¢ (YES 28/30¢) · Limit NO @ 73¢ · Effective entry band 60–90¢ favorite (72–73¢)
My probability (NO) 88% vs market-implied ~72% → edge ≈ +15¢, EV ≈ +20.5% (recorded for calibration only — did not set size)
Mispricing thesis. The market prices a 28% chance that this week's Hormuz transits exceed 100 — but as of the week's first day, the US is reinstating its naval blockade and has ordered new strikes, after IRGC missile attacks on three commercial ships Jul 6–8 collapsed traffic to a two-month low. Even if traffic merely persists at the crowd's own implied current run-rate (~13/day, per Kpler's Jul 8 count), the week totals ~91 — still under 100. YES therefore requires traffic to improve materially against an active blockade, i.e. a brand-new formal US–Iran deal within ~2–3 days plus an immediate insurer/tanker re-entry. The original June MOU took weeks of Islamabad-track negotiation to produce. The market is selling me that reversal at 28¢; I price it at ~8–12¢.
Evidence.
- Resolver data (IMF PortWatch, chokepoint6, n_total): full-closure floor Apr 20–May 31 averaged ~6/day (range 0–13); post-MOU recovery Jul 1–5 ran 42, 36, 34, 25, 34; normal baseline (Jul 2025) is ~71–119/day. Latest published datum is Jul 5 — the resolving week is a pure forecast for everyone.
- Jul 6–7 (night): IRGC fired missiles at the LNG tanker Al Rekayat and supertanker Wedyan; a third vessel was hit within 24h (Al Jazeera, Jul 7).
- Jul 8: 13 tankers crossed vs a 33/day prior-week average (Kpler, via CNBC Jul 9).
- Jul 10: Trump: talks continue but "the ceasefire is over" (Al Jazeera).
- Jul 13 (today, day 1 of the resolving week): traffic at a two-month low — lowest since May 25, a date on which PortWatch recorded 7 transits (US News); Trump orders the naval blockade of Iranian shipping reinstated and CENTCOM announces a new round of strikes (CNN, CBS). Iran still demands a toll regime for the strait, which the US and Oman call unacceptable — the core dispute is unresolved.
- Recovery-speed base rate: even after the signed Jun 17 MOU with a 60-day safe-passage guarantee, traffic took ~7 days to reach a single 52-transit day, and the best subsequent full week (Jun 24–30) totaled ~244 from an open strait. From today's ~7–13/day base, a >100 week needs the fastest diplomatic and commercial reversal of the entire crisis.
Resolver profile & priced risks (Stage 2.5). Resolver: IMF PortWatch daily chokepoint dataset — "the number of transit calls through the Strait of Hormuz as reported by the IMF PortWatch … summed from Jul 13 to Jul 19"; resolves no earlier than Tue Jul 21, 09:00, once complete data is available (rules_secondary). Mechanism is a pure arithmetic sum of a public feed — no discretion. Priced deductions: −2 pts PortWatch daily figures are revised and a near-threshold total could be nudged; −1 pt feed methodology/outage quirk forcing a delayed or alternative read. Evidence-based P(NO) ≈ 91 → final model_prob 88. Total deductions 3 pts — well under the 15-pt screen.
Tail risks (cleanest way to lose). Qatar-mediated shuttle diplomacy produces a signed de-escalation within ~48–72 hours and the queued tanker backlog surges through immediately (the June reopening shows pent-up queues can spike daily counts fast) — five strong days at 15–25+ clears 100. Secondary: my read of current traffic (from tanker-only Kpler counts) understates PortWatch totals, and actual current run-rate is ~15/day rather than ~7–13, in which case persistence alone gets close to the strike and any thaw crosses it.
Numbers 68 contracts × 73¢ = $49.64 cost · max payout $68.00 (+$18.36) · EV ≈ +$10.2
Liquidity (entry context) YES 28 bid / 30 ask · 68 contracts fillable at NO ≤73¢ (1 @ 72, 67 @ 73; next depth at 75¢+) · vol24h 6,587 · OI 5,308
Price history Market opened Jul 10; YES traded 21–79¢ on thin volume Jul 11, then settled 21–36¢ as the blockade news firmed (Jul 12–13). The NO side has cheapened ~2–4¢ into today's escalation headlines — entry is not chasing a move.
Cluster hormuz-no-fast-truce (sole member this run)
4. Recommended $1,000 portfolio
One pick cleared all screens. That is the finding, not a failure: this category's liquid slate this week is (a) war-path markets where specialist AIS watchers have already repriced the past week, (b) markets duplicating the 20 picks already live in the feed, (c) noise ladders, or (d) deep favorites offering ≤3¢ after resolver risk. v3 doctrine: cash is a position; low-conviction filler is what lost money in v1/v2.
| # | Ticker | Action | Limit ¢ | Contracts | Cost $ | Conviction (set size) | Band | Max payout $ | Edge ¢* | EV %* |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | KXHORMUZWEEKLY-26JUL19-T100 | BUY NO | 73 | 68 | 49.64 | HIGH (cap $150; sized to book depth) | 60–90¢ favorite | 68.00 | +15 | +20.5 |
| Cash reserve | 950.36 | 95.0% of capital — held, per the entry-band and conviction screens | ||||||||
| Total | 68 | 1,000.00 | Deployed $49.64 (5.0%) | 68.00 | +20.5 blended on deployed; dollar edge ≈ +$10.2 | |||||
Cluster exposure
| Cluster | Cost $ | % of capital | Cap (15%) | Status |
|---|---|---|---|---|
| hormuz-no-fast-truce | 49.64 | 5.0% | $150 | OK — well under cap |
Conviction exposure
| Tier | Cost deployed $ | % of capital | Per-pick cap |
|---|---|---|---|
| HIGH | 49.64 | 5.0% | $150 |
| MEDIUM | 0.00 | 0% | $70 |
| LOW | 0.00 | 0% | $30 |
Risk profile
- Worst case: a fast formal US–Iran deal plus an immediate tanker surge pushes the week above 100 → −$49.64 (−5.0% of capital).
- Best case / most likely (~88%): the week totals ≤100 → +$18.36 (+1.8% of capital, +37% on cost).
- Concentration: a single thesis carries all deployed capital, but at 5.0% of the book it cannot flip the run's sign materially; 95% of capital is in cash.
Execution notes
- Rest a NO limit at 73¢, good-till Wed Jul 15 close. Do not chase above 73¢ — the top-of-book is thin (68 contracts) and paying 75–80¢ erodes the band economics.
- Invalidation triggers (exit or stop adding): (1) a signed US–Iran de-escalation or restored MOU announced on or before ~Jul 15–16; (2) credible AIS reporting (Kpler/TankerTrackers) of daily transits back above ~20 by midweek; (3) PortWatch's next publication showing Jul 6–8 totals far above my ~15–35/day estimates, which would mean my current-traffic anchor is wrong.
- Watchlist: KXHORMUZWEEKLY-26JUL19-T75 BUY NO — modeled +30¢ edge but rejected by the coin-flip band rule (conviction MEDIUM). It becomes deployable only if PortWatch's Jul 6–12 publication (expected ~Jul 14–17) confirms a ≤10/day post-attack run-rate, hardening conviction to HIGH while the entry is still in-band.
5. What I rejected and why
All 14 Stage-2 survivors below are machine-logged in picks.json and shadow-tracked to settlement — if the rejects outperform the picks, the screen adds nothing, and we want to know that. Stage-3.5 entry-band cuts are marked ◆.
| Ticker | Side I'd have taken | Price ¢ | My prob % | Reject reason |
|---|---|---|---|---|
| KXHORMUZWEEKLY-26JUL19-T75 | NO | 51 | 81 | ◆ Coin-flip band (51¢) and conviction only MEDIUM — under the market's own AIS read of current traffic (~13/day), mere persistence resolves it YES; my lower traffic anchor is a 2× ambiguity. Modeled +30¢ edge logged for the shadow track; the band rule requires HIGH conviction here and I can't claim it. |
| KXHORMUZWEEKLY-26JUL19-T50 | NO | 30 | 31 | ◆ Sub-35¢ tail band and essentially zero modeled edge (31 vs 30). No trade either side — YES at 73¢ prices >50 above my 63% estimate. |
| KXHORMUZWEEKLY-26JUL12-T125 | NO | 27 | 50 | Resolving week already over; traders with live AIS repriced this continuously while PortWatch (the resolver) lags to Jul 5. My week estimate (~120–135) straddles the strike. Informational disadvantage — no defensible side. |
| KXHORMUZWEEKLY-26JUL12-T150 | NO | 81 | 85 | Same AIS disadvantage; price already collapsed 80→20¢ over five sessions on the thesis. A pre-blockade rush through the strait on Jul 6–7 (the crowd marks Jul 7 as the max day at 84%) could put the total right at 150 — I can't price that without their data. |
| KXHORMUZMAX-26JUL12-JUL07 | YES | 85 | — | Resolves on intra-week AIS facts already known to specialist traders; I formed no independent estimate. |
| KXHORMUZMAX-26JUL12-JUL06 | NO | 86 | — | Same — no independent information about which past day was the max. |
| KXMAXSHIPSHORMUZ-26JUL31-AL60 | YES | 6 | 8 | ◆ Sub-35¢ tail. Slight modeled YES edge (a new deal by ~Jul 24 + queue surge could print one 60+ day; June's best was 52), but the single per-run lottery slot is not used — tails went ~0-for in the resolved record. |
| KXHORMUZNORM-26MAR17-B260815 | NO | 96 | 95 | Fully-priced deep favorite (fails the ≥95¢ screen at the fill) and a near-duplicate of the live feed pick on the B260715 variant; 7-day MA >60 needs ~85% normalization within a month of an active blockade. |
| KXCLARITYVOTE-26JUL-AUG08 | YES | 66 | 70 | Price moved +24¢ in the last 48h on exactly this thesis (Senate returned Jul 13; merged draft due). Residual edge ~4¢ is inside my uncertainty: no cloture filed, Thune gave week 1 to the NDAA, and three disputes (ethics, §604, stablecoin yield) can still keep it off the floor entirely. Resolver is lenient (any recorded vote, incl. failed cloture) — logged for the shadow track. |
| KXHBUDGETRES-26JUN-26AUG01 | NO | 93 | 94 | Deep-favorite band. No FY2027 budget resolution exists, is noticed for markup, or is on any schedule (CRFB tracker: "TBD"; Budget Cmte's latest markup notice is May 2025) with ~10 legislative days left — but after a −2pt deduction for the resolver possibly counting a deeming resolution, the residual is ≤3¢. Not worth a 19-day lockup ("don't over-allocate chasing a few cents"). |
| KXTRUMPUFC-26AUG-DJT | YES | 14 | 25 | ◆ Sub-35¢ tail on a base-rate hunch (Trump attends several UFC cards a year; UFC 330 is a Saturday title card in swing-state Philadelphia near his Bedminster season). Zero primary evidence of attendance plans; price fell 22→13¢ on volume without visible news. Lottery slot not used. |
| KXTRUMPWORLDCUP-26-DJT-1 | YES | 92 | 94 | Duplicates the live feed pick KXTRUMPATTEND (final, Jul 19) — ≥1 match is priced identically (91/92) to the final-only market despite being strictly more probable, but the 1–2¢ structural gap is inside the spread. Re-picking would double the feed's exposure to one event. |
| KXLAGODAYS-26JUL-1 | NO | 92 | 95 | Complement of the live feed pick KXLAGODAYS-0 YES (93/94) — same underlying "no Mar-a-Lago trips in July" thesis; adding it duplicates exposure for ~3¢ in a >90 band. |
| KXKASHANNOUNCEOUT-26APR-AUG01 | NO | 94 | 96 | Tightly correlated with the live feed pick KXKASHOUT NO (announcement ⇒ departure market moves in lockstep); >90 band, ≤2¢ residual. |
6. Sources
- IMF PortWatch — resolution source; daily chokepoint data pulled via the public ArcGIS API (Daily_Chokepoints_Data, portid chokepoint6): Apr 20–May 31 avg ~6/day; Jul 1–5: 42/36/34/25/34; Jul 2025 baseline 71–119/day; latest datum Jul 5.
- Al Jazeera, Jul 7, 2026 — IRGC missile strikes on Al Rekayat and Wedyan; third vessel within 24h.
- CNBC, Jul 9, 2026 — Kpler: 13 tanker crossings Wed Jul 8 vs 33/day prior-week average.
- US News/Reuters, Jul 13, 2026 — traffic at two-month low (lowest since May 25) amid renewed US–Iran strikes.
- CNN live, Jul 13, 2026 — ceasefire disintegrates; Trump to reinstate blockade of Iranian ships in the strait.
- CBS News live — CENTCOM: new round of strikes ordered.
- Al Jazeera, Jul 10, 2026 — Trump: talks continue, ceasefire over; Qatar mediating.
- Axios, Jun 14, 2026 — ceasefire extension and strait reopening deal (context for the Jun 17 Islamabad Memorandum).
- CoinDesk, Jul 9, 2026; crypto.news; TechTimes, Jul 10, 2026; CryptoSlate — CLARITY Act status: no cloture filed, NDAA week, 7+ Dem crossovers needed.
- CRFB Appropriations Watch FY2027 — FY2027 budget resolution status "TBD" in both chambers; House Budget Committee markup notice (latest: May 2025, FY2025 reconciliation).
- UFC.com / Wikipedia — UFC 330: Aug 15, 2026, Xfinity Mobile Arena, Philadelphia (Makhachev vs Machado Garry).
Data: live Kalshi mirror (read-only SQL over trading_events / trading_markets / market_snapshots), Kalshi public trade API (top-of-book at ~2026-07-13T19:00Z), IMF PortWatch ArcGIS feed, and the cited press. All probabilities are subjective estimates; prediction-market contracts can and do resolve to zero. This v3 report's picks count toward the public track record and are cleared for auto-copy; the 14 logged rejects are shadow-tracked to settlement. Nothing here is investment advice.