Geopolitical oil price risk not a major driver of global macroeconomic fluctuations

There has been heightened interest in the economic effects of geopolitical risk in recent years. Oil supply risks have been widely considered among the top risks, reflecting uncertainty about OPEC quotas, access to Russian oil, widening conflict in the Middle East and disruptions of oil shipments.

As widely acknowledged in the academic literature, the mere possibility of geopolitical events disrupting oil production may raise oil price uncertainty and cause a recession.

Why do many economists expect oil price uncertainty to slow economic activity?

The origins of this perception can be traced to the 1980s. After two major oil price increases (in the early and late 1970s) were followed by major U.S. recessions, it was natural to conjecture that oil price increases cause recessions. However, by the same logic one would have expected major oil price declines to cause economic expansions.

This notion was put to a test in 1986, when efforts by OPEC to stabilize the price of oil collapsed and the oil price dropped as sharply as it had increased after the Iranian Revolution in 1979. Yet, the U.S. economy failed to expand in the wake of this oil price decline.

One interpretation of this evidence is that the relationship between the price of oil and the economy is linear. In this case, the modest response of the economy in 1986 implies that the effects of oil price shocks on the economy were also quite modest in 1979. In other words, the major recession in 1982 must have been caused by something other than the price of oil. A natural alternative explanation is the sharp monetary tightening that started in 1979 under Federal Reserve Chairman Paul Volcker.

An alternative interpretation of the same evidence is that the relationship between the price of oil and the economy is nonlinear, with changes in oil-price uncertainty accounting for the apparent asymmetry in the response to rising and falling oil prices. In this interpretation, the recession following the surge in the price of oil in 1979 was greatly worsened by a simultaneous increase in geopolitically driven oil price uncertainty, while in 1986 the stimulus implied by falling oil prices was largely offset by an increase in oil price uncertainty associated with the collapse of OPEC. 

The latter view has been supported by seemingly robust evidence from empirical work purporting to show large recessionary effects of higher oil price uncertainty. Typically, however, increases in oil price uncertainty in this literature have been taken as self-evident or inferred from statistical models under strong assumptions rather than directly observed. 

Oil price uncertainty relates to the volatility of the price of oil that cannot be predicted. A recent Dallas Fed Working Paper quantified for the first time the evolution of one-quarter-ahead oil price uncertainty since the early 1970s (Chart 1).

What drives oil price uncertainty?

It is important to stress that oil price uncertainty cannot be measured by changes in the variability of the growth rate of the price of oil. Nor can it be measured based on the number of times word combinations such as “oil price” and “uncertain” appear in news articles. Rather oil price uncertainty arises from lack of predictability. A natural definition of uncertainty is the variability in the future price of oil relative to the price predicted from currently available data.

As illustrated in Chart 1, surges in oil price uncertainty may be associated with geopolitical events but need not be driven by geopolitical events. For example, the largest spike in oil price uncertainty occurred in 2008 during the Global Financial Crisis and was clearly not driven by geopolitical events in oil markets. A good example of a geopolitical event causing a large spike in uncertainty was Iraq’s invasion of Kuwait in 1990. Yet a similar geopolitical event in 1980, the outbreak of the Iran-Iraq War, had little apparent effect on oil price uncertainty, indicating that geopolitical events need not be associated with higher oil price uncertainty.

In can also be difficult to isolate to what extent surges in oil price uncertainty were caused by geopolitical events as opposed to macroeconomic events. For example, the Iranian Revolution in 1979 coincided with an unprecedented shift in monetary policy under Volcker, and the drop in oil demand caused by the pandemic in early 2020 occurred at the same time as the Saudi price war. 

This evidence highlights the importance of modeling the endogenous determination of oil price uncertainty, macroeconomic uncertainty and economic fluctuations.

How to model linkages from geopolitical risk to economic fluctuations and the price of oil

We seek to understand the effects of geopolitical risk in a theoretical model of the global economy that is calibrated to U.S. macroeconomic, financial and global oil market data. Both oil prices and oil price uncertainty are simultaneously determined with the economy.

The calibrated model provides a good fit to the data, suggesting it is a useful laboratory for understanding the role of geopolitical oil price risk. One distinguishing feature of the model is its focus on one-sided risk.

We stress that what concerns firms and households is not predictive volatility so much as downside risk. For example, a consumer driving a car may be concerned about a geopolitical event disrupting oil production and causing the price of gasoline to reach $5 per gallon. That consumer will not be concerned, however, about the possibility of technological innovation increasing oil production to the point that the price of gasoline drops to 50 cents per gallon. In other words, the relevant risk for consumers is one-sided rather than two-sided.

In our model, uncertainty is driven mainly by downside risk. Downside risks to oil production, or for that matter to the economy, are subjective and hard to measure, but they can be calibrated. In the model, we consider two forms of downside risk that drive oil price uncertainty and macroeconomic uncertainty.

Macroeconomic disasters are modeled after events such as the Great Depression or the Global Financial Crisis. These disasters occur once every half century and on average last 2.5 years. Geopolitically driven oil production disasters are modeled after the historical oil production disasters in 1973–74, 1979 and 1990. They involve a 5 percent drop in global oil production that is expected to occur every 12.5 years. The focus of the study is how the anticipation of these disasters affects the economy rather than their realization.

How changes in the probability of a disaster impact oil markets and the global economy

Not only do the effects of anticipated disasters differ qualitatively from the effects of realized disasters as highlighted by our model, but also there are important differences between geopolitical oil price risk and macroeconomic risk.

For example, an increase in the probability of an oil production disaster causes higher oil price uncertainty and, hence, higher oil storage demand, reflected in a persistent increase in oil inventories and a higher oil price (see “Baseline” in Chart 2). This in turn causes a drop in investment and, hence, output, but little response in output uncertainty or consumption.

An increase in the probability of a growth disaster of the same size, representing increased macroeconomic disaster risk, in contrast, causes a much larger rise in output uncertainty, which is closely mirrored by higher oil price uncertainty. As in the case of an oil production disaster probability shock, there is a drop in investment and, hence, output, but the magnitudes are 10 times larger. As actual and expected demand for oil drops, there is a sharp drop in the oil price, and oil inventories persistently decline. Thus, the response of oil inventories and of the price of oil is of opposite sign than in the case of an oil disaster probability shock.

The effects of shocks to the downside risk to oil production are very different from those of more traditional stochastic volatility shocks to exogenous oil prices or exogenous oil production. It can be shown that the latter shocks have little effect on the economy because the resulting risk is not tilted to one side, but two-sided, highlighting the importance of modeling one-sided risks.

New insights on the role of geopolitical risk in oil markets

The model provides fresh perspectives on the role of geopolitical risk in oil markets. First, policymakers tend to be rightly concerned about geopolitical oil price uncertainty. Notwithstanding the attention geopolitical events in oil markets have attracted, however, we find that geopolitical oil price risk is unlikely to generate sizable recessionary effects.

Our analysis shows that even a large increase in the probability of an oil production disaster similar to disasters that have occurred historically does not have a large macroeconomic impact. For example, a 20-percentage-point increase in the probability of a major geopolitical oil disaster is only expected to lower output by 0.12 percent.

It is conceivable, of course, that households and firms at times were anticipating a longer lasting or larger oil production disaster than was ever realized. Notably, following the invasion of Kuwait in 1990, oil price uncertainty rose by more than implied by any of the observed oil production disasters in our model.

Such an increase in oil price uncertainty could be explained by anticipation of an oil production shortfall closer to 20 percent of global oil production, corresponding to the oil at risk from a war in the Persian Gulf (see “Larger disaster” in Chart 2). A disaster of the same magnitude as in the baseline, but expected to last 10 quarters rather than three quarters, in contrast, would not have much larger effects than in the baseline (see “Longer Disaster” in Chart 2).

Even when considering an unprecedented 20 percent drop in oil production, however, it would still take a large increase in the probability of such a disaster (or its realization) to cause a sizable recessionary impact. Such an event, if it were realized, would involve a 50 percent increase in the price of oil and 0.7 percent decline in output. 

Second, geopolitical oil price risk has not been an important driver of macroeconomic variability to date, intuitively because historically these risks have been rare and the geopolitical events in question are comparatively small.

Finally, in our model, more than half of the observed oil price uncertainty tends to be macroeconomic uncertainty. This suggests that previous empirical studies showing large effects of oil price uncertainty likely captured the effects of macroeconomic uncertainty rather than only geopolitical oil price uncertainty.

About the authors

Lutz Kilian is a vice president in the Research Department of the Federal Reserve Bank of Dallas.

Michael Plante is an assistant vice president in the Research Department at the Federal Reserve Bank of Dallas.

Alexander W. Richter is a vice president in the Research Department at the Federal Reserve Bank of Dallas.

The views expressed are those of the authors and should not be attributed to the Federal Reserve Bank of Dallas or the Federal Reserve System.