Kalshi Politics — Mispricing Audit

Report date 8 June 2026 · Category Politics · Horizon 45 days · Portfolio $1,000 · 322 markets screened → 3 picks
Bottom line. The politics board this week is dominated by a US–Israel–Iran war and a stalled nuclear-deal track, but those markets are either definitional traps (what counts as a "deal"?) or efficiently priced. The clean edges are three small, near-dated domestic markets: fade the TRUMPMAMDANI "meeting" that the market re-rated up on a misleading headline, buy MAMDANIEO on a base-rate gap, and take a thin lean against a LEAVESTARMER exit. No high-conviction names this week — sizing is deliberately modest and ~10% is held in reserve.

1 · How this was researched

  1. Screen. Queried the live Kalshi DB mirror for every active market whose event category = 'politics' closing within 45 days (322 markets). Joined the latest market_snapshots row per ticker and cut: 24h volume < 1,500; bid-ask > 5¢; already-priced (YES bid ≥ 95¢ / ask ≤ 5¢); and "what will X say"/mention noise. → 23 candidates.
  2. Rules. Pulled the full resolution text for the 14 most-liquid candidates. Roughly half the apparent edges died here on definitions (see §5).
  3. Tape. Pulled 14-day intraday history from market_snapshots to spot moves and check whether a thesis was already priced in.
  4. Book. Hit the live Kalshi public orderbook (/markets/{ticker}/orderbook) for each finalist to confirm I can size without paying ≥3¢ of slippage.
  5. News. Grounded every probability in primary sources — White House fact sheets, the House floor calendar, RealClearPolling, NYC Mayor's Office EO log, and named wire reporting. Sources in §6.

All prices are cents per contract (a contract pays $1.00 on YES). "NO @ 80" means buying the NO side for 80¢, paying out $1.00 if the event does not happen.

2 · Markets reviewed

TickerMarketYES bid/askVol 24hVerdict
KXTRUMPMAMDANI-26JUN09Trump & Mamdani meet in person before Jun 920 / 215,757PICK · BUY NO
KXMAMDANIEO-26JUN13-T0Mamdani signs >0 non-emergency EOs Jun 7–1312 / 133,395PICK · BUY YES
KXLEAVESTARMER-26JUL01Starmer leaves/announces leaving UK PM before Jul 113 / 144,636PICK · BUY NO
KXUSAIRANAGREEMENT-27-26JULUS agrees to new Iran nuclear deal before Jul 18 / 999,850PASS — definitional
KXHORMUZNORM-26MAR17…Hormuz transit 7-day MA > 60 before Jul 17 / 836,674PASS — thin/2-way
KXNEWTARIFFS-26JUN-JUL01Trump issues new/increased tariff action in June90 / 926,594WATCH — empty book
KXRECNCH-26-JUN10House passes reconciliation bill before Jun 1073 / 784,484WATCH — timing/5¢
KXFISAEXTEND-26MAY-JUN12FISA 702 reauthorization becomes law before Jun 1211 / 142,100PASS — knife-edge
KXTRUMPUFC-26JUL-DJTTrump attends UFC 329 (Jul 11, Las Vegas)22 / 256,098PASS — no signal
KXAPRPOTUS-26JUN12-39.0Trump RCP approval below 39.0 on Jun 126 / 73,019PASS — ~fair
KXHEGSETHOUT-26APR-JUL01Hegseth out as SecDef before Jul 14 / 65,256PASS — fair longshot
KXKASHOUT-26APR-JUL01Kash Patel out as FBI Director before Jul 19 / 92,748PASS — fair longshot
KXTRUTHSOCIAL-26JUN13-*Trump weekly Truth Social post-count buckets (9)~12kPASS — sum 105.5, no arb
KXLAGODAYS-26JUN-*Trump Mar-a-Lago trips in June (range set)~12kPASS — ~fair

Volume/price as of the latest 5-min snapshot, ~16:00Z 8 Jun 2026. The remaining ~9 screened candidates (UFC Freedom 250 celebrity-attendance markets, additional approval/Truth-Social legs) were celebrity-noise or duplicative and dropped.

3 · The picks

Ordered by conviction. Conviction reflects how defensible the edge is — source quality, how unambiguous the resolution rules are, and orderbook depth — not raw EV. There are no HIGH-conviction names this week.

Pick 1 — KXTRUMPMAMDANI-26JUN09 · BUY NO @ 0.80 · Medium conviction

"Will Donald Trump meet in person Zohran Mamdani before Jun 9, 2026?" · closes 2026-06-09 13:59Z (resolves in ~1 day)

Mispricing thesis. The only contact window left before the Jun 9 close is tonight's NBA Finals Game 3 at Madison Square Garden, which both men are slated to attend. A "both at the Knicks game" headline pushed YES back up toward ~20¢ today — but Mamdani has gone out of his way to say he will not be near Trump. The market is paying for an interaction the principal is actively avoiding.

Current: YES 20/21 → NO ask 80 True P(meeting): ~12% NO fair value: ~88¢ Edge: ~+8¢ Return at fill: ~10%

Evidence.

Cleanest way this loses: the two men cross paths at MSG and exchange a brief, photographed greeting that Kalshi rules an "in-person meeting." Two NYC-celebrity Knicks fans in the same building is a non-trivial path — hence MEDIUM, not HIGH, despite the strong sourcing.
Size: 550 contracts NO Limit: 0.81 (book fills ~0.805 avg) Cost: ~$443 Max payout: $550 EV: ~+$44

Liquidity: ~287 contracts at NO 80, then ~1,360 more at NO 81 (resting YES bids of $57 @20¢ and $259 @19¢). 550 lots fill inside 1¢ of slippage.

Pick 2 — KXMAMDANIEO-26JUN13-T0 · BUY YES @ 0.13 · Medium conviction

"Will Mamdani sign above 0 executive orders between Jun 7–13, 2026?" (non-emergency) · closes 2026-06-14 13:59Z

Mispricing thesis. This is a base-rate market. Mayor Mamdani has been an unusually prolific executive-order signer, yet the contract prices "at least one non-emergency EO this week" at only ~13%. His empirical weekly hit rate is closer to ~23–26%.

Current: YES 12/13 Empirical weekly P: ~23–26% True P (discounted): ~19–20% Edge: ~+6¢ Return at fill: ~46%

Evidence.

Cleanest way this loses: his output is bursty (5 EOs in January, then a zero in March), and he just fired one on Jun 1 — so Jun 7–13 could simply be a quiet week with no policy roll-out, which is the modal outcome (most individual weeks are zeros). The edge is the base-rate gap, not a known catalyst; size accordingly.
Size: 1,800 contracts YES Limit: 0.13 Cost: ~$234 Max payout: $1,800 EV: ~+$108

Liquidity: ~2,080 contracts resting at YES 13 (NO bid $271 @87¢). 1,800 lots fill at the limit with no slippage.

Pick 3 — KXLEAVESTARMER-26JUL01 · BUY NO @ 0.87 · Low conviction

"Will Keir Starmer leave / announce leaving as UK PM before Jul 1, 2026?" · closes 2026-07-01 03:59Z

Mispricing thesis. A genuine Labour leadership crisis is under way, but the contract resolves YES only if Starmer actually leaves or formally announces his intention to leave within a 3-week window — and he is publicly refusing to do either. This is a thin, honest lean, not a strong edge.

Current: YES 13/14 → NO ask 87 True P(leave/announce): ~10% NO fair value: ~90¢ Edge: ~+3¢ Return at fill: ~3.4%

Evidence.

Cleanest way this loses: UK leadership crises move fast — a triggered confidence vote or a wave of cabinet resignations could produce a departure announcement well inside three weeks. With only ~3¢ of edge, this is a small position, not a core holding.
Size: 250 contracts NO Limit: 0.87 Cost: ~$218 Max payout: $250 EV: ~+$7

Liquidity: very deep — thousands of contracts resting at YES 12–13 ($24.5k @12¢). No slippage concern.

4 · Recommended $1,000 portfolio

PickSideLimitContractsCostMax payoutTrue PEV ¢/ctEV $Conv.
TRUMPMAMDANINO0.805504435500.88+8+44Med
MAMDANIEOYES0.131,8002341,8000.19+6+108Med
LEAVESTARMERNO0.872502182500.90+3+7Low
TOTAL8952,600+159
Deployed: ~$895 Cash reserve: ~$105 (10.5%) Total max payout: $2,600 Aggregate EV: ~+$159 Blended EV on deployed: ~+17.8%

Risk profile

Execution notes

5 · What I rejected and why

6 · Sources

Data sources. Live Kalshi DB mirror (trading_markets, trading_events, market_snapshots) for the screen, prices, volume and 14-day tape; Kalshi public API orderbooks for depth; web search/fetch of the primary sources above for priors. Snapshot prices as of ~16:00Z, 8 Jun 2026.

Disclaimer. Not investment advice. Every probability here is a subjective estimate, and prediction-market contracts can resolve to zero — you can lose 100% of any stake. Edges are small and depend on resolution rules being read as written; read each market's official rules before trading. Prices and books move continuously and may already differ from those quoted.