Oil Price Shocks

Definition

An oil price shock is an unexpected movement in the (real) price of crude oil, and the empirical literature is concerned with decomposing such movements into structurally distinct driving forces — chiefly oil supply disruptions versus oil demand shifts — and tracing their dynamic effects on macroeconomic aggregates (real GDP growth, inflation) via impulse responses. The central modern insight, developed by Lutz Kilian, is that "the" oil price is an endogenous equilibrium object: not all oil price increases are alike, and their macroeconomic consequences depend on what caused them.

Key Ideas

Endogeneity of the oil price. The real price of oil is jointly determined by global production, real activity, and precautionary demand; treating oil-price increases as an exogenous supply-side tax on the economy is generally invalid.
Supply shocks explain surprisingly little. Exogenous OPEC production shortfalls (wars, political disturbances) account for only a small fraction of crisis-period oil-price increases — a demand/precautionary story is needed for most episodes (Kilian 2008).
Event-based exogenous supply measure. A counterfactual construction of what OPEC production would have been absent an exogenous political event — benchmarked against uninvolved producers — yields a shock series that (unlike quantitative dummies) allows delayed, long-lasting, time-varying, and even sign-reversing responses.
Weak instruments. Using exogenous production shortfalls to instrument the oil price in macro projections suffers a weak-instrument problem; a direct OLS projection of the macro aggregate on current/lagged shocks is a cleaner way to estimate the effect of the exogenous component (Kilian 2008).
Episodic, not uniform, macro effects. Exogenous supply shocks produce a delayed GDP-growth drop and inflation spike, but their average contribution to U.S. macro fluctuations since the 1970s is modest — large only in particular episodes (e.g. the 1990/91 Gulf War).

How It Works

Two complementary empirical strategies dominate. (1) Narrative / event-based measurement builds an exogenous oil-supply-shock series from historical accounts of production disruptions, then projects macro aggregates on it by OLS (Kilian 2008). (2) Structural VAR models of the global crude-oil market decompose oil-price movements into supply, aggregate-demand, and oil-specific (precautionary) demand shocks, and read off impulse responses and historical decompositions. The two agree on the headline conclusion that demand, not exogenous supply, drives most oil-price variation.

The three-shock structural VAR (Kilian 2009)

The canonical decomposition uses a monthly VAR in $z_t=(\Delta prod_t, rea_t, rpo_t)'$ — percent change in world crude production, an index of global real economic activity, and the (log) real oil price — with 24 lags, identified by a recursive (Cholesky) ordering of $A_0^{-1}$ :

Oil supply shock — innovation to global production; oil supply is assumed not to respond within the month to demand shocks (a vertical short-run supply curve).
Aggregate demand shock — innovation to global real activity not explained by supply, reflecting the global business cycle.
Oil-specific (precautionary) demand shock — innovation to the real oil price not explained by the first two; captures a shift in the conditional variance (uncertainty over future shortfalls), i.e. convenience-yield / inventory-insurance demand.

The three shocks have distinct dynamics — a precautionary shock hits oil prices immediately and persistently, an aggregate-demand shock with a delay, a supply disruption only weakly and transiently — and a historical decomposition attributes the 1979 and post-2003 price surges mainly to demand, not physical supply. The real-activity index itself is a Kilian contribution: detrended, CPI-deflated dry-cargo ocean freight rates as an exchange-rate-free monthly proxy for the global industrial-commodity cycle. Because the composition of shocks shifts over time, reduced-form regressions of macro aggregates on the oil price are inherently unstable.

Why It Matters

Monetary/fiscal response. The appropriate policy reaction to an oil-price increase depends on its source: a supply-driven increase is stagflationary, whereas a demand-driven increase reflects a booming global economy.
Forecasting and risk. Correctly attributing oil-price movements is essential for inflation forecasting and for pricing oil-related macro risk.
Methodological caution. Reframes decades of "oil shocks cause recessions" evidence that had implicitly treated oil prices as exogenous.

Open Questions

Identification of the structural oil-market shocks (recursive vs. sign-restriction vs. informative-prior schemes) remains contested — see Bayesian SVAR Identification and the Baumeister-Hamilton critique.
How much precautionary/uncertainty demand (as opposed to flow demand) drives oil prices, and how to measure it.