Part V · Commodities, macro and geopolitics · M12 · week 7 · 18 h

Geopolitics and event studies

Event-study econometrics, the geopolitical risk index, scenario trees with Bayesian updating, options as geopolitical hedges, transmission channels into Indian prices, and seven case studies from 1990 to 2026.

Start reading · Learning outcomes

Builds on M02 Statistics of returnsM04 Systematic researchM08 Option pricingM09 Volatility tradingM10 CommoditiesM11 Macro

Prerequisites M02 (returns, fat tails), M04 (hypothesis testing, multiple testing), M08–M09 (option payoffs, implied volatility), M10 (crude and gold), M11 (rates, rupee, flows). Maths: OLS in matrix form and its sampling distribution, the $t$ and $\chi^2$ distributions, CLT, Bayes' theorem. Time: ~18 hours (notes 10 h, problems 5 h, code 3 h). Primary texts: MacKinlay (1997), "Event Studies in Economics and Finance"; Campbell, Lo and MacKinlay, The Econometrics of Financial Markets (1997), ch. 4; Caldara and Iacoviello (2022), "Measuring Geopolitical Risk".

Labels: [V] means checked against a primary or reputable source (EIA, IMF, central banks, journal pages, major outlets) on or near 6 Nov 2026. [S] means secondary or aggregator data only (Wikipedia, Yahoo Finance price history, press aggregators), so treat it as indicative. Many 2026 facts are from news and aggregators because primary sources have not yet published final data. They are labelled accordingly.

Learning outcomes

Set up a market-model event study: choose estimation and event windows, estimate normal returns, and compute abnormal returns (AR) and cumulative abnormal returns (CAR).
Derive the exact finite-sample variance of a CAR, including estimation error, and prove the standardised CAR is $t$ -distributed under the null.
Derive the cross-sectional test statistics (J1, Patell, Boehmer–Musumeci–Poulsen) and show by simulation when each has correct size.
State how the Caldara–Iacoviello GPR index is built, what its threats and acts sub-indices measure, and what its regressions find.
Update scenario probabilities on news with Bayes' rule in probability, odds and log-odds form, and compute expected values, the value of information and decision-tree policies.
Explain why traders hedge geopolitical risk with options, and when a hedge has positive expected value under one's own probabilities.
Trace a geopolitical shock to Indian assets through oil chokepoints, risk premia, sanctions, trade, capital flows and the rupee.
Analyse seven episodes (1990, 2022, 2023–24, Apr 2025, May 2025, 2026, Nov 2026) with dated, sourced numbers, and say what each teaches.

Sections ≈ 1 h 55 min of study

Foundation 3 sections · ≈ 20 min
Core 7 sections · ≈ 1 h 15 min
Advanced 2 sections · ≈ 20 min

Learning outcomes Foundation 5′
12.1 Event-study design Core 5′
12.2 Abnormal returns and their exact distribution Core 15′
Lab Event study on real data: Nifty, Brent, gold and the rupee around a shock
12.3 Aggregating across events and firms Advanced 10′
Figure Simulated event study: CAR, J1, Patell and BMP
12.4 Measuring geopolitical risk: the GPR index Core 10′
12.5 Scenario analysis: decision trees and Bayesian updating Core 10′
Figure Bayesian scenario updating: worked example 12.2
12.6 Asymmetric payoffs: convexity and options as geopolitical hedges Advanced 10′
Figure A put as a geopolitical hedge: worked example 12.3
12.7 Transmission channels: how geopolitics reaches Indian prices Foundation 10′
12.8 Case studies Core 25′
Summary of key results Foundation 5′
Readings and resources Core 5′
Facts that change Core 5′