Geo Scout — News Sentiment Trading

While Theta Sniper profits from time decay (a mechanical edge), Geo Scout profits from information asymmetry — finding markets where the current price doesn't reflect information that you have access to but the broader market hasn't processed yet.

For Pakistani traders, this is a structural advantage. The regions you understand best — South Asia, the Middle East, Pakistan's domestic politics — are exactly the regions where Western-dominated Polymarket consistently mispricies events.

The Geo Scout Thesis

Polymarket's user base is predominantly US and European. When a market asks "Will Pakistan-India diplomatic relations improve in Q2 2026?", the price is set largely by traders who:

Read English-language news with a 6–12 hour delay on South Asian events
Lack understanding of the actual political dynamics between the two governments
Cannot access Urdu, Hindi, or Pashto sources where early signals appear
Underestimate or overestimate risk based on Western media framing

You, as a Pakistani trader who monitors Dawn, Geo News, ARY News, and local Twitter/X in real time, have a structural information advantage that is not available to the majority of the market.

The GEO_KEYWORDS Scanning System

The first step is filtering: Polymarket has thousands of active markets. We only want ones where our information edge applies.

python

# strategies/geo_scout.py

# Tier 1: Pakistani/South Asian markets where our edge is highest
PAKISTAN_KEYWORDS = [
    'pakistan', 'sbp', 'imf pakistan', 'cpec', 'psl',
    'babar azam', 'shaheen afridi', 'lahore', 'karachi',
    'imran khan', 'nawaz sharif', 'pti'
]

# Tier 2: Regional markets where we have above-average understanding
SOUTH_ASIA_KEYWORDS = [
    'india', 'india-pakistan', 'kashmir', 'bangladesh',
    'sri lanka', 'afghanistan', 'saarc'
]

# Tier 3: Global geopolitical markets (general advantage, weaker edge)
GLOBAL_GEO_KEYWORDS = [
    'iran', 'israel', 'ukraine', 'russia', 'china', 'taiwan',
    'nato', 'election', 'tariff', 'trade war', 'oil', 'opec'
]

# Assign edge weights by tier
KEYWORD_TIERS = {
    'tier_1': {'keywords': PAKISTAN_KEYWORDS, 'edge_multiplier': 1.5},
    'tier_2': {'keywords': SOUTH_ASIA_KEYWORDS, 'edge_multiplier': 1.2},
    'tier_3': {'keywords': GLOBAL_GEO_KEYWORDS, 'edge_multiplier': 1.0},
}

def find_geo_markets(all_markets: list) -> list:
    """
    Filter markets by geo keywords and assign edge tier.
    Returns markets with tier classification.
    """
    geo_markets = []

    for market in all_markets:
        question = market.get('question', '').lower()
        best_tier = None
        matching_keywords = []
        primary_theme = None

        for tier_name, tier_data in KEYWORD_TIERS.items():
            found_kws = [kw for kw in tier_data['keywords'] if kw in question]
            if found_kws:
                if best_tier is None:  # First (highest) tier match wins
                    best_tier = tier_name
                    matching_keywords = found_kws
                    primary_theme = found_kws[0]

        if best_tier:
            market['_geo_keywords'] = matching_keywords
            market['_theme'] = primary_theme
            market['_tier'] = best_tier
            market['_edge_multiplier'] = KEYWORD_TIERS[best_tier]['edge_multiplier']
            geo_markets.append(market)

    return sorted(geo_markets, key=lambda x: x['_edge_multiplier'], reverse=True)

Theme-Based Position Limiting

Geo Scout's biggest risk: correlated positions. If you hold 5 markets that all resolve based on Pakistan-India relations, and a major diplomatic incident occurs, all 5 resolve against you simultaneously.

The protection: limit total exposure per theme.

python

class GeoPositionTracker:
    """Track and limit positions per geopolitical theme."""

    def __init__(self, db_connection):
        self.db = db_connection
        self.THEME_LIMITS = {
            'pakistan': 2,      # Max 2 open positions on Pakistan-specific markets
            'india': 2,         # Max 2 on India-specific markets
            'iran': 2,          # Max 2 on Iran-specific markets
            'election': 3,      # Elections are less correlated, allow 3
            'default': 2        # Default for all other themes
        }
        self.TOTAL_GEO_CAP = 6  # Maximum total open Geo Scout positions

    def can_enter_position(self, theme: str) -> tuple[bool, str]:
        """
        Check if a new position on this theme is allowed.
        Returns (allowed: bool, reason: str)
        """
        # Check theme-specific limit
        theme_positions = self.db.count_open_positions_by_theme(theme)
        theme_limit = self.THEME_LIMITS.get(theme, self.THEME_LIMITS['default'])

        if theme_positions >= theme_limit:
            return False, f"Theme limit reached: {theme_positions}/{theme_limit} for '{theme}'"

        # Check total Geo Scout position cap
        total_geo_positions = self.db.count_open_positions_by_strategy('geo_scout')
        if total_geo_positions >= self.TOTAL_GEO_CAP:
            return False, f"Total Geo Scout cap reached: {total_geo_positions}/{self.TOTAL_GEO_CAP}"

        return True, "Position allowed"

The Geo Scout Signal Pipeline

code

GEO SCOUT SIGNAL FLOW

News source (Dawn, Geo TV, Reuters, Bloomberg)
         │
         ▼
RSS feed collector (fetches every 15 minutes)
         │
         ▼
GEO_KEYWORDS filter
(does this headline match any of our watched themes?)
         │
         ▼ (if match found)
Gemini Flash triage
"Is this news BULLISH or BEARISH for outcome X on Polymarket?
Confidence: 0-1"
         │
         ▼ (if confidence > 0.55)
Claude Sonnet deep analysis
"Assess the true probability of this market resolving YES
given this news. Explain your reasoning. Account for
Pakistani political context."
         │
         ▼
Ensemble aggregation
(combine all signals into final probability estimate)
         │
         ▼
Edge calculation
(ensemble estimate - current market price = edge)
         │
         ▼ (if edge > 0.12 and theme limit not exceeded)
EXECUTE TRADE

Pakistani Information Advantage by Category

code

INFORMATION EDGE ASSESSMENT FOR PAKISTANI TRADERS

HIGH EDGE (structural advantage over Western traders)
┌──────────────────────────────────────────────────────┐
│ SBP Monetary Policy decisions                        │
│  ▶ Dawn/sbp.org.pk has real-time press releases      │
│  ▶ Pakistani economists tweet before English media   │
│  ▶ IMF EFF constraints well-understood locally       │
│  Edge window: 15-45 minutes ahead of price move      │
├──────────────────────────────────────────────────────┤
│ Pakistan cricket (PSL, Test, T20 matches)            │
│  ▶ Local coaches/analysts on Urdu Twitter            │
│  ▶ Pitch reports from PakPassion, PCB.com.pk         │
│  ▶ Player injury news reaches English media slowly   │
│  Edge window: 30-90 minutes ahead of price move      │
├──────────────────────────────────────────────────────┤
│ Pakistan domestic political events                   │
│  ▶ Urdu media covers PTI/PMLN developments faster   │
│  ▶ Local political analysts have better context      │
│  Edge window: 1-4 hours ahead of English media       │
└──────────────────────────────────────────────────────┘

MODERATE EDGE
┌──────────────────────────────────────────────────────┐
│ India-Pakistan diplomatic developments               │
│  ▶ Both sides of the narrative accessible in Urdu   │
│  ▶ Still better than Western traders' single-source │
│  Edge: Moderate — Indian market also has edge        │
└──────────────────────────────────────────────────────┘

LOW EDGE (no structural advantage)
┌──────────────────────────────────────────────────────┐
│ US politics, EU events, Western markets              │
│  ▶ Same information access as global traders         │
│  ▶ No local context advantage                        │
│  Action: Trade only with strong AI ensemble signal   │
└──────────────────────────────────────────────────────┘

Practice Lab

Exercise 1: Set up the GEO_KEYWORDS dictionary with at least 15 keywords relevant to your information advantage. Divide them into tiers: (a) markets where you have genuine, real-time Urdu-language information advantage, (b) markets where you have moderate advantage, (c) general global markets. Run the find_geo_markets function against a list of 20 fictional market questions and verify it correctly classifies each.

Exercise 2: Build the GeoPositionTracker class with a mock database (use a Python dictionary instead of a real DB). Test: (a) add a position on theme "pakistan", (b) add another on "pakistan" (should succeed), (c) try adding a third on "pakistan" — it should be blocked. Print the reason at each step.

Exercise 3: Monitor Dawn.com, Geo.tv, and ARY News for 3 days. Every time you see a headline that matches one of your Geo Scout keywords, find the corresponding Polymarket market (if it exists). Note: (a) the time of the news, (b) the current market price, (c) your estimate of the true probability given the news, (d) what the price was 2 hours later. This measures your real information edge window.

Key Takeaways

Geo Scout's edge is information asymmetry — you have access to information the majority of the Western-dominated Polymarket user base doesn't process in real time
Keyword tiering matters: Pakistani domestic markets have the highest edge, South Asian regional markets moderate edge, global markets need stronger AI signal to compensate
Theme-based position limits prevent correlated exposure: 5 open positions on Pakistan-related markets become 5 simultaneous losses if a major domestic event reverses the narrative
The information edge window is measured in minutes to hours, not days — the bot must process news and act before the broader market reprices
Trading Geo Scout markets where you don't have genuine local knowledge is just speculation; the strategy only works when your cultural and language access genuinely puts you ahead of the Western traders setting the price

Pakistan Case Study: The Kartarpur Corridor Trade

Background: In January 2026, a Polymarket market existed: "Will the Kartarpur Corridor remain open for all of 2026?" Current YES price: 52 cents.

Hamid's Geo Scout analysis (a Lahore-based political science graduate now running a trading bot):

His Urdu-language news sources showed:

Both Pakistani and Indian governments had publicly reaffirmed the corridor's importance as a "humanitarian gesture above politics" in December 2025
No military exercises or troop movements near Kartarpur in 90 days
Pakistani and Indian media both had overwhelmingly positive framing of the corridor as a "soft diplomacy symbol"

Western traders were pricing YES at 52% because they associated "Pakistan-India relations" with conflict risk — a generic framing that ignored the specific political insulation of this corridor.

Hamid's ensemble estimate: 81% probability YES.

Edge: 81% - 52% = 29 cents — strong Tier 2 signal.

He entered $75 in YES shares. Over 30 days, the market drifted to 78 cents with no significant Pakistan-India incidents.

He exited at 78 cents. P&L: (0.78 - 0.52) × ($75 / 0.52) = +$37.50 = PKR 10,500 profit on a $75 trade.

Hamid's lesson: "My edge wasn't that I predicted the future better. My edge was that I understood the political context — that Kartarpur specifically was protected from broader tensions by both governments' domestic political calculations. Western traders saw 'Pakistan-India' and discounted it to 52%. I saw a specific, protected humanitarian gesture and priced it at 81%. That's the Geo Scout edge in a single example."

5.3 — Geo Scout — News Sentiment Se Trading