Exergy International · Political Risk Resilience

Abqaiq oil-facility strike · 14 Sep 2019

Case study · event class: Infrastructure Failure

On 14 September 2019, drone strikes on Saudi Arabia’s Abqaiq facility knocked out roughly 5% of global oil supply in a single day. The question worth asking: would the Political Risk Resilience (PRR) Framework have given you an edge?

You can’t forecast a shock like this. But you can carry an expectation into it. The PRR Framework reads how each firm’s value has moved across the infrastructure-failure event class — every other event of this kind. Leave this event out, and the class still tells you which firms are wired to gain and which to lose. That’s the read you’d have walked in with.^†

Before the event · what the class led you to expect, and what happened

event expectation vs outcome — Horizontal: what PRR’s read of the infrastructure-failure class led you to expect, set from the *other* events. Vertical: what each firm’s value actually did on the event.

The class flagged Sibanye · NMDC · Impala to gain and Petra Diamonds to lose — and on the day, that is broadly how it split.

The edge wasn’t calling the event — it was carrying a grounded expectation into it, set from how these firms metabolize this entire class of shock.

Why it matters — the next shock

That is the information gain. PRR can’t tell you when the next infrastructure-failure shock lands or how hard. It can tell you what to expect: substitutes and spare-capacity names tend to hold or gain, while the names most exposed take the hit. You carry that read into the next one.

The full class behind the read

This event is one observation. The expectation rests on the whole infrastructure-failure class — the structure the single event is drawn from.

class quadrant — The full infrastructure-failure class: each firm by its observed value response against whether PRR’s read aligns.

The evidence behind the read — for the analyst

Firm	Class expectation (ex-event)	On the day	Class alignment
Sibanye	+4.7%	+47.5%	strong (ρ +0.72)
NMDC	+8.3%	+26.3%	moderate (ρ +0.59)
Impala	+7.6%	+17.5%	moderate (ρ +0.53)
Century Al	+0.8%	-13.7%	strong (ρ -0.90)

† Method (beta, stylized example). A recognizable event chosen to illustrate the edge — not a comprehensive backtest. “Class expectation” = the firm’s mean value response across all other infrastructure-failure events (the focal event left out), 30-day window. “On the day” = the firm’s 30-day market-adjusted (abnormal) return around the event. “Class alignment” (ρ) = whether PRR’s firm read corresponds to which events moved the firm most. Firm value via cumulative abnormal return (CAR) in beta; more metrics planned. Exploratory / candidate-stage — association, not causation. PRR tells you what to expect; it does not forecast.