Politics Desk Strategy Spec

Polymarket as Sharp Anchor + Government Data Pipelines

Date: 2026-03-28


OVERVIEW

The Politics desk covers 65+ recurring Kalshi series across presidential approval, legislation, executive actions, elections, and government operations. There is no single "Pinnacle" for politics. Polymarket is the closest equivalent — real-money, deep liquidity, sophisticated international traders. We build a synthetic sharp line from Polymarket (primary) + Metaculus + Manifold + PredictIt + our own models.

Edge comes from two sources:

  1. Cross-market arbitrage — Polymarket-to-Kalshi price divergence on the same event
  2. Data-anchored contracts — Government data (Congress.gov, Federal Register, CBP) feeds probability models faster than the Kalshi crowd updates

SHARP ANCHOR STRATEGY

Polymarket is the political Pinnacle. Highest volume, most sophisticated traders, covers most Kalshi political markets.

Consensus Probability Formula

consensus_prob = (0.45 * polymarket) + (0.25 * metaculus) + (0.20 * manifold) + (0.10 * model)

For data-anchored contracts (bills, border, EOs): model weight increases to 0.40, others decrease proportionally.

Edge Detection

edge = consensus_prob - kalshi_implied_prob
trade_threshold = 0.08 (8 cents)
min_sources = 2 (at least 2 independent sources must agree)

CORE EDGE DETECTION

Cross-Market Arbitrage

Kalshi and Polymarket price the same political events. When Polymarket moves and Kalshi lags, that's the trade. Historical lag: 15-60 minutes depending on contract liquidity.

Data-Anchored Contracts

Some contracts have objective, trackable underlying data:

When the data moves and the Kalshi price hasn't, that's edge.

Position Sizing

1/4 Kelly for all political bets. Hard caps:


APPROVED MODELS (from panel ruling)

Approval Momentum Model

Legislative Readiness Score

Track bills through stages: introduced → committee → floor → passed_one → passed_both → enrolled → signed. Each stage increases readiness score. Cross-reference with Kalshi "will bill X pass" contracts.

Executive Surprise Model

EO/proclamation frequency modeled as Poisson process with contextual multipliers. When frequency deviates from baseline, "EO count" contracts are mispriced.


KEY DATA SOURCES

Source URL Auth Cost Purpose
Polymarket CLOB clob.polymarket.com/markets None Free Primary sharp anchor
Polymarket Gamma gamma-api.polymarket.com/markets None Free Market metadata
PredictIt predictit.org/api/marketdata/all/ None Free Cross-validation
Manifold Manifold API None Free Cross-validation
Metaculus Metaculus API None Free Forecaster consensus
Congress.gov api.congress.gov/v3/bill Free API key Free Bill pipeline
Federal Register federalregister.gov/api/v1/documents None Free EOs, proclamations
CBP Encounters cbp.gov/newsroom/stats/ None Free Border data
RCP Polling realclearpolling.com None Free Approval average
538 Approval projects.fivethirtyeight.com/polls/ None Free Approval model
GDELT 2.0 data.gdeltproject.org/gdeltv2/ None Free Geopolitical events
WH Schedule whitehouse.gov/schedule/ None Free Lid prediction
MARAD Alerts maritime.dot.gov/msci/msci-alerts None Free Geopolitical signal (Phase 3)

TRADEABILITY TIERS

Tier 1 — Recurring, Data-Anchored (highest priority)

Contract Type ~Series Anchor Data
Presidential Approval KXAPRPOTUS, KX538APPROVE RCP + 538 polling averages
EO Count KXTRUMPEO Federal Register
Bill Passage Various Congress.gov pipeline
Border Encounters KXBORDERENCOUNTER CBP monthly data

Tier 2 — Recurring, Market-Consensus

Contract Type ~Series Anchor
Election markets Various Polymarket primary
Cabinet/appointment Various Polymarket + news
Government shutdown Various Congress bill pipeline

Tier 3 — One-offs (mostly skip)

~3,158 one-off contracts. Stale price scan only.


COLLECTION SCHEDULE

Time (ET) Job Frequency
6:00 AM RCP + 538 approval scrape Daily
6:15 AM Congress.gov bill pipeline Daily
6:30 AM Federal Register EOs/proclamations Daily
Every 30 min Kalshi politics prices Continuous
Every 30 min Polymarket politics Continuous
Every 30 min PredictIt contracts Continuous
7:15 AM, 1 PM, 7 PM Politics consensus computation 3x daily
1st of month CBP border encounters Monthly
Weekly Sun 10 PM Pollster house effect update Weekly

WHEN NOT TO BET

Condition Reason
Net edge < 8 cents Below threshold (higher than sports)
Kalshi volume < $200 Too illiquid
< 2 independent sources agree Not enough consensus
Breaking news cycle (first 2 hours) Markets are chaotic, let them settle
Contract expires > 60 days out Too much uncertainty
One-off contract with no data anchor No model, no edge

PANEL RULING REFERENCE

Full ruling: /home/ubuntu/edgeclaw/results/panel-results/politics-data-final-ruling.md Panel: Opus (judge) + Sonnet + Grok 4.2 Reasoning + Gemini Pro 3.1 Date: 2026-03-26, Grade: A


TODO: Upgrade Kalshi REST to WebSocket Feed

Status: NOT BUILT — add when this desk goes live for execution

Current state: All Kalshi data (prices, order books, trades) is fetched via REST API polling on cron schedules. This is fine for edge detection and monitoring, but NOT sufficient for live trade execution.

Why WebSocket matters:

Note: Sports markets use RFQ (Request for Quote) so the visible order book is usually empty — but the trade feed still matters for freshness signals and the WebSocket is required for order submission. Weather and politics markets DO have real visible order books where this upgrade is even more critical.

Added 2026-03-29 — upgrade REST to WS when desk moves to live execution

Source: ~/edgeclaw/results/spec-panel/politics-strategy.md