Geopolitical Market Impact

Geopolitical Market Impact

Overview

Geopolitical events create asymmetric, non-linear market shocks. The challenge is: (1) classifying events by type and severity, (2) estimating impact direction and magnitude from historical analogues, and (3) acting before the crowd prices it in — or avoiding false alarms.

Core principle: Markets often front-run geopolitical risk. The best entry is frequently on rumour, not confirmation. Study buy the rumour, sell the news carefully.

When to Use

• Classifying incoming geopolitical news for trading impact
• Searching for historical precedents to estimate price response
• Building event-driven trading signals from news feeds
• Estimating sanction/war/election impact on specific assets
• Assessing second-order effects (e.g., oil → airline stocks)

Event Classification Framework

from enum import Enum
from dataclasses import dataclass
from typing import Optional

class EventType(Enum):
    WAR_ESCALATION    = "war_escalation"
    WAR_DE_ESCALATION = "war_de_escalation"
    SANCTIONS_NEW     = "sanctions_new"
    SANCTIONS_LIFTED  = "sanctions_lifted"
    ELECTION_RESULT   = "election_result"
    POLICY_CHANGE     = "policy_change"      # rates, tariffs, regulation
    NATURAL_DISASTER  = "natural_disaster"
    SUPPLY_DISRUPTION = "supply_disruption"  # Suez, pipelines, ports
    COUP_INSTABILITY  = "coup_instability"
    DIPLOMATIC_CRISIS = "diplomatic_crisis"

class Severity(Enum):
    LOW    = 1   # localised, limited economic ties
    MEDIUM = 2   # regional spillover, moderate trade impact
    HIGH   = 3   # global supply chain, major economy involved
    EXTREME = 4  # nuclear risk, G7 direct involvement

@dataclass
class GeopoliticalEvent:
    event_type: EventType
    severity: Severity
    countries_involved: list[str]
    affected_commodities: list[str]   # e.g. ['crude_oil', 'wheat', 'natural_gas']
    affected_currencies: list[str]    # e.g. ['RUB', 'EUR', 'USD']
    description: str
    source_reliability: float         # 0-1, Reuters=0.9, Twitter=0.3

Asset Impact Matrix

# Historical directional impact by event type and asset class
# Values: typical directional bias (+ = up, - = down, ? = uncertain)
IMPACT_MATRIX = {
    EventType.WAR_ESCALATION: {
        'gold':          +1,    # safe haven
        'usd':           +1,    # safe haven flows
        'oil':           +1,    # supply disruption risk
        'natural_gas':   +1,
        'wheat':         +1,    # Russia/Ukraine = major exporters
        'equities':      -1,    # risk-off
        'btc':           -1,    # risk-off (initially)
        'bonds':         +1,    # safe haven
        'eur':           -1,    # if European theatre
    },
    EventType.SANCTIONS_NEW: {
        'affected_country_currency': -1,
        'oil':           +1,    # if sanctioned country is oil exporter
        'gold':          +1,
        'usd':           +1,
    },
    EventType.ELECTION_RESULT: {
        # Highly context-dependent — use historical_precedents() function
    },
    EventType.SUPPLY_DISRUPTION: {
        'affected_commodity': +1,
        'substitutes':   +1,
        'consumers':     -1,    # companies dependent on commodity
    },
}

def estimate_impact(event: GeopoliticalEvent, asset: str) -> dict:
    """Estimate directional impact and magnitude."""
    matrix = IMPACT_MATRIX.get(event.event_type, {})
    direction = matrix.get(asset, 0)

    # Severity multiplier
    magnitude_pct = {
        Severity.LOW: 0.5,
        Severity.MEDIUM: 2.0,
        Severity.HIGH: 5.0,
        Severity.EXTREME: 15.0,
    }[event.severity]

    return {
        'direction': direction,
        'estimated_move_pct': direction * magnitude_pct,
        'confidence': event.source_reliability,
        'note': 'Directional estimate only; timing and confirmation critical',
    }

Historical Precedent Database

HISTORICAL_PRECEDENTS = [
    {
        'event': 'Russia invades Ukraine (Feb 2022)',
        'type': EventType.WAR_ESCALATION,
        'impacts': {
            'crude_oil':    +28,   # % move in 3 months
            'wheat':        +60,
            'natural_gas':  +200,  # EU benchmark
            'gold':         +8,
            'eur_usd':      -6,
            'rub_usd':      -50,   # initial, then recovered on capital controls
            'spx':          -12,   # in 3 months
        },
        'time_to_peak': '3-6 weeks',
        'reversal': 'Partial within 3 months as supply routes adapted',
    },
    {
        'event': 'US-China tariff escalation (2018-2019)',
        'type': EventType.POLICY_CHANGE,
        'impacts': {
            'cny_usd':      -8,
            'spx':          -20,   # peak drawdown
            'soybeans':     -15,
            'gold':         +8,
        },
        'time_to_peak': '6-12 months',
    },
    {
        'event': 'Iran nuclear deal JCPOA exit (2018)',
        'type': EventType.SANCTIONS_NEW,
        'impacts': {
            'crude_oil':    +20,
            'irr_usd':      -70,   # Iranian rial
        },
        'time_to_peak': '6 months',
    },
    {
        'event': 'Brexit referendum (Jun 2016)',
        'type': EventType.ELECTION_RESULT,
        'impacts': {
            'gbp_usd':      -10,   # day after
            'ftse_100':     -8,    # day after (recovered quickly as GBP offset)
            'gold':         +5,
            'eurostoxx':    -8,
        },
        'time_to_peak': '1-2 days for initial shock',
    },
]

def find_precedents(event: GeopoliticalEvent, top_n: int = 3) -> list[dict]:
    """Match current event to historical precedents by type."""
    matches = [p for p in HISTORICAL_PRECEDENTS
               if p['type'] == event.event_type]
    return matches[:top_n]

Second-Order Effects

# Commodity → sector impact chains
SECOND_ORDER_CHAINS = {
    'crude_oil_up': {
        'airlines':       -1,   # fuel costs
        'shipping':       -1,
        'petrochemicals': +1,   # feedstock price up → margin down for consumers
        'oil_majors':     +1,
        'ev_sector':      +1,   # accelerated adoption
        'defense':        +1,   # increased military spending
    },
    'natural_gas_up': {
        'utilities':      -1,   # margin compression
        'lng_exporters':  +1,   # US LNG benefits
        'fertilisers':    -1,   # gas = feedstock for nitrogen fertilisers
        'food_producers': -1,
    },
    'wheat_up': {
        'food_companies': -1,
        'agri_tech':      +1,
        'emerging_markets': -1, # food import dependent nations: Egypt, Lebanon
    },
}

Timing Model

def geopolitical_signal_timing(event_severity: Severity) -> dict:
    """
    Markets typically price geopolitical events in phases.
    Returns suggested trade timing parameters.
    """
    timing_map = {
        Severity.LOW: {
            'initial_spike_duration': '1-4 hours',
            'peak_impact': '1-3 days',
            'mean_reversion': '1-2 weeks',
            'action': 'Wait for initial spike, fade if no escalation',
        },
        Severity.MEDIUM: {
            'initial_spike_duration': '4-24 hours',
            'peak_impact': '1-2 weeks',
            'mean_reversion': '4-8 weeks',
            'action': 'Trade with trend during peak, watch for plateau',
        },
        Severity.HIGH: {
            'initial_spike_duration': '1-3 days',
            'peak_impact': '4-12 weeks',
            'mean_reversion': '3-6 months',
            'action': 'Structural position; manage with trailing stops',
        },
        Severity.EXTREME: {
            'initial_spike_duration': '3-7 days',
            'peak_impact': 'Months to years',
            'mean_reversion': 'Uncertain',
            'action': 'Risk-off immediately; reassess when situation stabilises',
        },
    }
    return timing_map[event_severity]

Quick Reference — Reliable Safe Havens

Common Mistakes

1. Chasing the initial spike — first reaction is often wrong; wait for second-day confirmation
2. Ignoring resolution probability — a "warning shot" sanctions vs. full embargo = very different impact
3. Forgetting sanctions evasion — Russia/Iran showed sanctions rarely achieve stated economic goals
4. Linear extrapolation — market prices in a lot quickly; impact often peaks before situation resolves
5. Missing second-order effects — the wheat price impact on Egyptian pound is often more tradeable than wheat itself
6. Source reliability — Telegram/Twitter geopolitical news has 50%+ false alarm rate; weight accordingly

Anti-Patterns