Does The Dollar Drive Oil Prices?

|
Includes: BNO, DBO, DNO, DTO, OIL, OILK, OILX, OLEM, OLO, SCO, SZO, UCO, UDN, USDU, USL, USO, UUP
by: Elliott Gue

Summary

The dollar is often used to explain changes in the price of crude oil over time.

There is a negative correlation (statistically significant inverse relationship) between changes in the value of the dollar and the price of oil.

Based on data over the past two decades, changes in the value of the dollar, only explain around 10% of moves in WTI and 6% in Brent.

Readers often ask me if the value of the dollar relative to other key global currencies is a major factor driving oil prices.

The short answer: The dollar and the price of oil are correlated but it’s not as important as many investors, and pundits in the mainstream financial media, seem to believe.

Since crude is a global commodity priced in US dollars, it’s logical to think that when the value of the dollar declines, you’d need more dollars to buy the same 42-gallon barrel of oil. In other words, the dollar and crude oil should be inversely correlated, moving in opposite directions.

Let’s start with a quick look at the historic data in chart form:

Source: Bloomberg

This chart shows the price of Brent crude oil (the key international oil benchmark) plotted against the US Dollar Index, which tracks the value of the dollar against a basket of major world currencies. I’ve inverted the scale for the Dollar Index to make the relationship, and the trends between the two series, easier to compare.

It’s clear there’s a broad relationship between oil and the dollar apparent in the data back to 1999. For example, Brent oil prices jumped over 400% between 2003 and their 2008 highs as the value of the dollar index fell about 29% over the same time period.

Similarly, as oil prices fell sharply from over $100/bbl in mid-2014 to under $30/bbl in early 2016, the value of the dollar index jumped more than 30%.

However, the magnitude of moves in oil prices dwarf those in the US dollar and there are plenty of examples of periods when the relationship has broken down. For example, the Dollar Index currently trades at around 95.25, up around 3.7% over the past year; yet, Brent oil prices have soared 46.6% over the same period from around $54 to nearly $79 per barrel.

First, the fact that oil and the dollar have both increased in value over the past year is not consistent with the inverse correlation relationship we’d anticipate. And, second, it’s tough to argue that a 3.7% change in the value of the dollar can explain a move that’s more than 12 times larger in percentage terms for crude.

To look at this more scientifically, check out my table:

Index Dollar Index WTI Crude Oil S&P 500 Brent Crude Oil S&P Energy Alerian MLP
Dollar Index 1.00 -0.32 -0.32 -0.25 -0.38 -0.23
WTI Crude Oil -0.32 1.00 0.22 0.92 0.55 0.31
S&P 500 -0.32 0.22 1.00 0.21 0.60 0.40
Brent Crude Oil -0.25 0.92 0.21 1.00 0.51 0.30
S&P Energy -0.38 0.55 0.60 0.51 1.00 0.48
Alerian MLP -0.23 0.31 0.40 0.30 0.48 1.00

Source: Bloomberg

This table shows the correlation coefficients for Brent, WTI, the S&P 500 Index, the S&P 500 Energy Index, the Alerian MLP Index and the Dollar Index over the past 30 years based on monthly data. All the correlations on this chart are significant at better than a 99% level of confidence, meaning the relationships are not likely due solely to chance.

A correlation coefficient of 1 means perfect positive correlation - the two assets move in the same direction and in lockstep. The correlation coefficients of 1 moving down the diagonal of my table simply shows that an asset is perfectly correlated to itself.

Similarly, a correlation coefficient of 0 indicates no relationship between the prices of two indexes or assets and a correlation coefficient of -1.0 indicates a perfect inverse correlation.

Thus, this table shows that both West Texas Intermediate and Brent crude oil are negatively correlated to the US dollar (negative signs on the correlation coefficient) – this is the inverse relationship that we anticipated. However, the correlation for Brent is only -0.25 and for WTI it’s just -0.32. That suggests the relationship is weak.

Indeed, the positive correlation between oil prices and the S&P 500 is almost as large in absolute terms as the negative correlation to the US dollar. The correlation between stocks and oil may be the result of the “risk on, risk off” moves over the past two decades where all risky assets – including stocks, commodities and high-yield bonds – tend to move in the same direction in response to macroeconomic forces such as the credit crunch in 2008-09, the European credit crisis in 2010 – 2012 and, more recently, headlines about “trade wars.”

Also, note that the negative correlation between the S&P 500 and the US Dollar Index is at least as strong as that between the dollar and crude oil.

Meanwhile the negative correlation between energy stocks – as measured by the S&P 500 Energy Index and the Alerian MLP Index – and the dollar is roughly the same in magnitude as for the dollar and oil itself.

Of course, the positive relationship between energy stocks and the broader market is much stronger than for the dollar and energy stocks – after all, energy stocks are still equities and will tend to follow the broader market, at least directionally, over the long haul.

Let’s dig a little deeper and further assess the strength of the relationship between oil, energy stocks and the US dollar:

Index Dollar Index WTI Crude Oil S&P 500 Brent Crude Oil S&P Energy Alerian MLP
Dollar Index 1.00 0.10 0.10 0.14 0.14 0.05
WTI Crude Oil 0.10 1.00 0.05 0.85 0.29 0.09
S&P 500 0.10 0.05 1.00 0.04 0.35 0.16
Brent Crude Oil 0.06 0.85 0.04 1.00 0.25 0.09
S&P Energy 0.14 0.29 0.35 0.25 1.00 0.23
Alerian MLP 0.05 0.09 0.16 0.09 0.23 1.00

Source: Bloomberg

This simple table is based on a regression of monthly changes in the price of oil against monthly changes in the value of the Dollar Index (and other assets), over the past 20 years.

Simply put, if we plot a trendline through these datapoints, the R-squared measures the tightness of fit for that trendline to the actual data, with an R-squared of 100% indicating a perfect fit for the trendline and an R-squared of 0 indicating no fit whatsoever.

In other words, R-squared answers the question: “How much of monthly changes in the price of oil can be explained by monthly changes in the value of the dollar index.”

As you can see, the answer is not much.

The R-squared for the dollar-WTI relationship is just 0.103 (10.3%) and for the value of Brent, it’s 6.1%. While statistically significant, that’s a very weak relationship.

Meanwhile, as you can see from my table, moves in the S&P 500 explain about 35% of the monthly moves in the S&P 500 Energy Index and 16% of moves in the Alerian MLP Index.

Bottom line: While there’s a meaningful relationship between oil and the dollar, trying to forecast moves in oil and energy stocks by focusing on the value of the currency or changes in the value of the currency is unlikely to provide a satisfactory outcome.

Disclosure: I am/we are long USO.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.