Published as part of the ECB Economic Bulletin, Issue 7/2024.
In recent years rising oil prices have often coincided with a strengthening of the US dollar – which has potentially intensified inflation dynamics in the euro area. Historically, there has been no clear link between oil prices and the US dollar (Chart A, panel a). In the period after the global financial crisis, the correlation tended to be negative. This co-movement was most likely the result of specific shocks – in particular shifts in global risk aversion – which sent oil prices and the US dollar in opposite directions (Chart A, panel b), rather than the reflection of a causal relationship.[1] However, recent studies suggest that the emergence of the United States as an oil exporter has been a factor in rendering the correlation consistently positive.[2] As crude oil is mainly traded in US dollars, a systematically positive co-movement would imply that the price of oil in local currency is more volatile than the dollar price of oil. This could strengthen the inflationary impact of oil shocks in net oil importers such as the euro area. The empirical models presented in this box show that the co-movement observed still seems to be largely the result of specific shocks that have steered both variables in the same direction rather than the reflection of a structural change in the link between oil prices and the US dollar.
Chart A
Developments in and correlation between oil prices and the US dollar
a) Oil prices and the US dollar | b) Correlation between oil prices and the US dollar |
---|---|
Sources: J.P. Morgan, US Energy Information Administration and ECB staff calculations.
Notes: The latest observations are for August 2024. Panel a): “USD REER index” is the US dollar real narrow effective exchange rate index. “Oil price” is the spot USD price per barrel of West Texas Intermediate deflated by the US consumer price index. Panel b): 24 and 48 months rolling correlations are computed for month-on-month changes in the variables.
The link between oil prices and the US dollar might have changed when the United States started to produce and export oil on a large scale. US crude oil production has more than doubled in the last 15 years (Chart B, panel a), driven by the shale oil boom and the lifting of the crude oil export ban by US Congress in 2015. As a result, the United States became a net exporter of oil in late 2019.[3] There are two main ways in which this might have altered the link between oil prices and the US dollar. First, it may have changed the effect of oil supply shocks on the US terms of trade – which are linked to exchange rates via the trade balance. The terms of trade for commodity products and oil prices, which had been closely correlated, started to gradually decouple in the United States after 2015, and the correlation switched sign when the country became a net exporter (Chart B, panel b). Second, the surge in US oil production may have changed the response of domestic aggregate demand to oil supply shocks. For instance, higher oil prices could make investment in the oil sector more attractive, supporting domestic demand and thus causing the US dollar to strengthen.
Chart B
US crude oil production, oil prices and US commodity terms of trade
a) US crude oil production | b) Oil prices and US commodity terms of trade |
---|---|
Sources: International Energy Agency, US Energy Information Administration, Citi and ECB staff calculations.
Notes: The latest observations are for August 2024. Panel b): “Oil price” refers to the West Texas Intermediate oil spot price deflated by the US consumer price index. “Terms of trade” refers to the Citi Commodity Terms of Trade Index.
The empirical results presented in this box indicate that the structural change in the US oil market has not been sufficient to turn the oil-dollar correlation systematically positive. A monthly local projections model based on estimated oil supply shocks is used to analyse time variation in the link between oil prices and the US dollar. The estimation period is split into three episodes, each capturing a different step in the transformation of the United States into a net exporter of oil.[4]
- In the period from 1989 to 1999, when US oil production was low and the country was still a net importer of oil, supply shocks that caused oil prices to rise led to a deterioration in the terms of trade and a weakening of the US dollar – which was as to be expected for an oil-importing country (Chart C, left-hand column).
- In the period from 2000 to 2010, when oil production started to ramp up but the country was still mostly an importer, oil supply shocks began to support domestic industrial production – albeit not to any significant extent. The effect on the terms of trade remained negative, however, as the United States was still a net importer of oil. This implied that the response of the US dollar to a shock to oil supply was less negative – insignificant even – during this period (Chart C, middle column).
- In the most recent period, from 2011 to 2023, when the country produced significant amounts of crude oil and became a net exporter, the terms of trade no longer deteriorated after oil supply shocks caused prices to rise. At the same time, the boost to industrial production increased in line with stronger domestic oil production. But overall, according to the model, these changes were not forceful enough to significantly lift the US dollar and create a systematic positive link between the two variables, as the dollar’s response to oil supply shocks during this period was negative (Chart C, right-hand column).
Chart C
Response of US variables to oil supply shocks over time
There are two main factors potentially explaining the absence of a stronger positive effect on the US dollar. First, US net oil exports as a share of total exports are still low compared with countries such as Norway or Australia that have “commodity currencies” which clearly appreciate following oil supply shocks (Chart D). Second, it could be because higher oil prices dampen global trade, which is dollar-intensive and thus has a negative impact on the US dollar.[5]
Chart D
Exchange rate response to oil supply shocks and net exports of commodities
The positive correlation still seems be explained by a combination of shocks rather than a structural change in the link between oil prices and the US dollar. A daily Bayesian vector autoregression (BVAR) model, which jointly identifies the main drivers of movements in oil prices and the US dollar, can be used to assess the importance of oil shocks compared with other drivers of US dollar dynamics.[6] According to this model, the positive co-movement between oil prices and the US dollar is mostly the result of a specific constellation of shocks that occurred in the past which happened to send oil prices and the US dollar in the same direction. This can be illustrated by decomposing oil prices and the US dollar into their main drivers during a recent episode characterised by a persistently high positive correlation (Chart E).
- From June 2021 to June 2022, as the US economy emerged from the COVID-19 shock, the Federal Reserve System embarked on its monetary tightening cycle and Russia launched its invasion of Ukraine, much of the US dollar’s strength was explained by the Federal Reserve’s tighter monetary policy, together with a stronger than expected US economy (Chart E, panel a). At the same time, oil prices rose sharply, mostly as a result of disrupted supply related to geopolitical risks (Chart E, panel b).
Chart E
Historical decomposition of the US dollar and oil prices – selected episodes
a) US dollar | b) Oil prices |
---|---|
Source: ECB staff calculations.
Notes: The charts show the historical decomposition of the cyclical component of the nominal broad US dollar and oil prices using a daily BVAR model with sign and narrative restrictions. The model is estimated over the period 2015-24.
Overall, these results indicate that the positive co-movement often observed between oil prices and the US dollar is not likely to be systematic. The correlation still reflects a specific constellation of shocks. A strengthening US dollar might at times reinforce the inflationary impact of higher oil prices in oil-importing countries, while a weakening dollar might dampen this impact. Over time, as the United States becomes a larger net exporter of oil, the link between oil shocks and the US dollar could continue to move further into positive territory.
Past studies have shown that spikes in global risk aversion largely explained the negative correlation around the time of the global financial crisis, see Fratzscher, M., Schneider, D. and Van Robays, I., “Oil prices, exchange rates and asset prices”, Working Paper Series, No 1689, ECB, 2014.
Hofmann, B., Igna, D. and Rees, D., “The changing nexus between commodity prices and the dollar: causes and implications”, BIS Bulletin, No 74, Bank for International Settlements, 2023, and Rees, D., “Commodity prices and the US Dollar”, BIS Working Papers, No 1083, Bank for International Settlements, 2023. This positive correlation is not a new development, however, as the correlation was also positive at times before 2007, when the United States was still a net importer of oil.
According to US Energy Information Administration data, the United States is a net oil exporter when petroleum products are included but a net importer when considering crude oil alone.
The local projection model regresses the variables shown in Chart C on the oil supply shocks estimated in Baumeister, C. and Hamilton, J.D., “Structural Interpretation of Vector Autoregressions with Incomplete Identification: Revisiting the Role of Oil Supply and Demand Shocks”, American Economic Review, Vol. 109, No 5, May 2019, pp. 1873-1910. The model uses six lags and controls for economic activity, inflation and monetary policy. The sample used ranges from January 1989 to March 2023.
See Boz, E., Casas, C., Georgiadis, G., Gopinath, G., Le Mezo, H., Mehl, A. and Nguyen, T., “Patterns of invoicing currency in global trade: New evidence”, Journal of International Economics, Elsevier, Vol. 136, 2022.
The model considers the following variables: US NEER, oil prices, commodity terms of trade, US equities (excluding the oil sector) and US two-year yields. Shocks are identified using sign and narrative restrictions. The sample considered is for the period 2015-24 to account for the structural change observed in the US economy.