Link Between Large Oil Price Changes and GDP: It's Not What You Think

With oil’s recent moves, there has been talk of potentially adding government regulation to rein in some types of oil speculation. The thundering chorus seems to pick up volume whenever oil spikes up rather than when oil spikes down which makes sense after oil’s recent run from the lows a few months back. But before any hasty decisions are made, it might make sense to see what happens to GDP growth during periods of large changes in oil. I used a 20% change in the price of oil as an area of focus. Do higher oil prices lead to lower GDP growth, which is the common thinking? After looking at the data I am not so sure that this is truly the case. After reading below feel free to drop your comments along. This will allow the reader to settle this debate.

First, I used the Energy Information Administration US Government and Statistics web site to gather date for changes in quarterly oil prices. I used the Cushing Oklahoma Spot WTI prices and used the last day of each quarter as my price points. I also used the US Department of Commerce Bureau to gather the data for GDP. For GDP I used the chained 2000 data gathering quarterly changes in GDP which are annualized rates. There were 92 quarterly periods in total beginning with Q2 1986 and ending with Q1 2009. I began with Q2 1986 because there was no daily price data for 1985 for the WTI price. During these 92 periods, oil rose on average 3.60% and GDP rose on average 2.69%.

I then culled the data for any time that oil moved up or down on a quarter to quarter closing basis by at least a 20% change. It turns out there were 19 periods when oil moved by at least 20% during any quarterly period. Of the 19, 12 saw oil moving up by 20% or more, and 7 saw oil moving down by 20% or more. During these 12 periods of increasing oil prices, GDP went up 11 times and once was unchanged. Oil averaged an increase of 37.84% while GDP rose an average of 3.42%. The 7 times that oil dropped by 20% or more, GDP rose 3 times but fell 4 times. During these 7 periods, the average change for oil was -17.79% and the average GDP increased by 0.25%. Looking again at the complete 92 periods, oil on average increased by 3.60% and GDP on average rose 2.69%, but during times when oil rose by 20% or more, GDP rose 3.42% and never once did GDP fall. During the 7 oil price drops, GDP on average increased by an average of 0.25%. Higher oil prices and higher GDP increases seem to go together, and lower oil and lower changes in GDP as well.

Here is the data for the 19 quarterly periods when oil moved by 20% or more from Q2 1986 through Q1 2009 and its corresponding GDP change :

Qtr	Oil %	GDP %
1986q2	24.88	1.6
1986q4	21.97	2.0
1988q4	28.43	5.4
1990q3	131.85	0.0
1994q2	31.06	5.3
1999q1	37.23	3.4
1999q3	26.95	4.8
2000q2	20.77	6.4
2002q1	31.31	2.7
2004q3	34.24	3.6
2005q1	27.56	3.0
2008q2	37.84	2.8
Average	37.84	3.42
Qtr	Oil %	GDP %
1990q4	-27.95	-3.0
1991q1	-31.07	-2.0
1993q4	-24.2	5.5
1997q1	-21.43	3.1
1998q4	-25.02	6.2
2008q3	-28.05	-0.5
2008q4	-55.71	-6.3
Average	-17.79	0.25

What does this all mean? When oil prices change by 20% or more on a quarterly basis, GDP increases at a larger than average pace of 3.42% and that when oil prices fall by 20% ore during a quarterly basis, GDP rises on average by an anemic 0.25%. Now why does this happen? Does the change in oil signify that demand for other products and services is strong and therefore acts as a rationing process thereby forestalling shortages before they actually occur? Maybe, because that’s what higher prices do though. They force the user to cut back and therefore avoid a potential shortage. The same holds true for lower prices, the glut on the market forces the producer to curtail production. Or is the common thinking that a spike in oil will lead to an economic slowdown. There may be other data that show this but at least during this period that did not seem to be the case.

I ran this exercise because a large move up in the price of oil may not be the contractionary force on GDP that most assume. Large price drops may not lead to an increase in GDP but may be signaling a contraction or slowing of the economy. There are times when we all want the price of energy to stop going higher, but those forces eventually correct as the economy slows. Higher oil prices do not seem to correspond with a slowing GDP (at least not during this period). Before regulations are added to oil trading, links between oil price changes and GDP growth should be studied carefully.

Authored by Tom Henderson, Strategist at JBH Capital.

There are no disclosures to make.

Comments

[I am not a ...:]

Perhaps we who use a lot fuel have to work harder to maintain our lust for pleasure at the wheels of our respective vehicles ?
This mass effort of the populace is reflected as a rise in GDP?
I don't know... but it sounds feasible,certainly a lot more realistic than the G20's moves to increase oil costs to limit consumption and thus save the planet warming by 2 Degrees C.

Jul 10 08:24 AM

[Robert Mart...:]

Tom: Would it make sense to lag the GDP data by one quarter? This would show how rising oil prices hurt growth going forward (if that's the thesis you are testing). You may have to play with the lag period, which may be anywhere from 3 to 9 months. I would also consider using global GDP (or G8 data) to capture world demand.

My experience is that oil and GDP are positively correlated for the obvious reason: Both signal global demand. The marginal barrel of demand sets the oil price at the bottom of the cycle, while the marginal barrel of supply sets the price at the top of the cycle. Supply & demand for oil are both highly inelastic in the short run, so price volatility is exaggerated by tiny changes in either variable.

Jul 10 08:26 AM

[john s. gordon:]

in 2008 we had a large runup in oil $ carefully timed by the speculators & hedgie funds to match the summer driving season. consumers having been skinned cut back on all other purchases during the ensuing 12 months which is the effect we are seeing now.
> jack

Jul 10 08:28 AM

[paulsjj:]

Hows the GDP doing now after oil hit $140.00 brl last year?

Jul 10 08:46 AM

[A Barrel Full:]

I think that using a lagging analysis makes a lot of sense. Short term increases in oil price are largerly ignored. Its only when high prices persist, that consumer behaviour statrs to change.


On Jul 10 08:26 AM Robert Martorana wrote:

> Tom: Would it make sense to lag the GDP data by one quarter? This
> would show how rising oil prices hurt growth going forward (if that's
> the thesis you are testing). You may have to play with the lag period,
> which may be anywhere from 3 to 9 months. I would also consider using
> global GDP (or G8 data) to capture world demand.
>
> My experience is that oil and GDP are positively correlated for the
> obvious reason: Both signal global demand. The marginal barrel of
> demand sets the oil price at the bottom of the cycle, while the marginal
> barrel of supply sets the price at the top of the cycle. Supply &
> demand for oil are both highly inelastic in the short run, so price
> volatility is exaggerated by tiny changes in either variable. <br/>

Jul 10 08:48 AM

[The Inflect...:]

Your logic makes sense. The more we use something, the higher the cost. This will lead to a cutback in demand and potentially to alternative energy sources.

Today is Nikola Tesla's birthday, the great electrical engineer who worked in Europe and in America with Edison. I studied some of his work in engineering school in college. Maybe some great mind like Tesla's will come along to help solve some of our energy problems, until then supply and demand will have to suffice.

Thanks for the comment.

Tom Henderson


On Jul 10 08:24 AM I am not a number.. wrote:

> Perhaps we who use a lot fuel have to work harder to maintain our
> lust for pleasure at the wheels of our respective vehicles ?
> This mass effort of the populace is reflected as a rise in GDP?<br/>I
> don't know... but it sounds feasible,certainly a lot more realistic
> than the G20's moves to increase oil costs to limit consumption and
> thus save the planet warming by 2 Degrees C.

Jul 10 09:06 AM

[The Inflect...:]

Excellent comment. I thought of looking at the potential of oil's effect with a lag and may pursue this further.

The inelastic comment is vital in this argument too because if there is only so many barrles of oil (x) available, if demand takes off and we reach (x-1) barrles, then the price roars ahead.

Thanks for the comment.

Tom


On Jul 10 08:26 AM Robert Martorana wrote:

> Tom: Would it make sense to lag the GDP data by one quarter? This
> would show how rising oil prices hurt growth going forward (if that's
> the thesis you are testing). You may have to play with the lag period,
> which may be anywhere from 3 to 9 months. I would also consider using
> global GDP (or G8 data) to capture world demand.
>
> My experience is that oil and GDP are positively correlated for the
> obvious reason: Both signal global demand. The marginal barrel of
> demand sets the oil price at the bottom of the cycle, while the marginal
> barrel of supply sets the price at the top of the cycle. Supply &amp;
> demand for oil are both highly inelastic in the short run, so price
> volatility is exaggerated by tiny changes in either variable. <br/>

Jul 10 09:11 AM

[The Inflect...:]

Thanks you for the comment.

I was shocked to see oil take off to $140 last summer. In my mind it made no sense because the economy was in a tailspin already. But it really came down quickly and violently.

I did not like seeing oil up at $140 last summer and prefer the recent lows in the $30s. (for the record, nor did I participate in last year's runnup except by telling those around me that it would not last) .

There may be something else to this argument - that the US economy is not as effected by a change in oil as much as it used to be. I clearly remember the 1970's and although there were other factros affecting the US economy oil seemed to be one of the largest culprits. The punch from oil does not carry the same power (thankfully ) on the US economy as it did years ago.

Thanks for the comment.

Tom


On Jul 10 08:28 AM john s. gordon wrote:

> in 2008 we had a large runup in oil $ carefully timed by the speculators
> &amp; hedgie funds to match the summer driving season. consumers
> having been skinned cut back on all other purchases during the ensuing
> 12 months which is the effect we are seeing now.

Jul 10 09:18 AM

[The Inflect...:]

Touche! Good point. But, did oil really push us into the recession we are in? I don't believe so and as I said above the economic horizon had already darkened and to me in the summer of 2008 it felt like we were already in an economic hurricane. Actually, the official start of the recession was declared to be in late (November / December 2007).

Did $140 oil hurt the economy at all? Absolutely. It may have pushed the US economy ever so slightly further into recession. But the $30 / oil since has not seemd to help much as GDP has had its 2 worst quarters during the collapse in oil from $140 to $30.

Thanks for the comment.

Tom Henderson


On Jul 10 08:46 AM paulsjj wrote:

> Hows the GDP doing now after oil hit $140.00 brl last year?

Jul 10 09:24 AM

[The Inflect...:]

Great comment. I may take a look at the lagging effect that oil may or may not have on GDP.

Thanks

Tom Henderson


On Jul 10 08:48 AM A Barrel Full wrote:

> I think that using a lagging analysis makes a lot of sense. Short
> term increases in oil price are largerly ignored. Its only when high
> prices persist, that consumer behaviour statrs to change.

Jul 10 09:26 AM

[change is t...:]

The overall analysis seems sound. However, if oil's price is both the source of growth and the reason for recession.... oil's price lead's and lags will feedback into the GDP numbers. Speculators causing a short term shock to costs may have a larger affect than can be quantified - it will depend on individual companies' ability to manage balance sheets with access to credit. Thinly capitalized companies will die with a severe shock. Adaquately capitalized finished goods producers at either full capacity, undercapacity and dealing with a credit contraction will all be affected by oil costs. The true effect oil has on GDP, and GDP on oil, gets into Black swan territory if you analyze the last 18 months like the previous decade. Never before has a oil shock been preceded by or followed by a credit crunch. Including a black swan event into your data stream about oil and GDP, that arguably includes a feedback bias, may leave you with only garbage.

Jul 10 03:15 PM

[G. L. Turner:]

In the initial rise in oil prices, in normal times - meaning that the GS CIF and other manipulation is not the main driver - the price of oil is reflecting the expansion of the economy, as it should. However, as one commentor suggested, and as I think too, a more appropriate look-see would be to examine the next 2/3 Qs that followed, and see what actually happens. Also, maybe 20% is the pinnacle as to where oil can rise in a 'growth tandem' with the economy, oil prices helping to support GDP - "to a point!"

My question then, what is the effect on GDP when oil continues to rise above the inflection point of 20%.

In my business, I see a rather large knee-jerk reaction to a 50 cent rise in gasoline prices, and have rolled up and down since mid-2005, following Bush's OIL COMPANY mergers, that have pushed us to a doulbling to tripling price tier ever since the mergers. Many business start to feel such pinches, gradually...however, 50 cents in gas does not necessarily translate, or stem from a 20% rise in oil....

Jul 11 05:49 PM

[The Inflect...:]

I see, so the 20% may be the point up to which the economy can grow without adverse affect. And it might be possible that once you go beyond a 20% oil increase that GDP could then start feeling it.

Good point, I will have to look into that.

Thanks for the comment.

Tom


On Jul 11 05:49 PM G. L. Turner wrote:

> In the initial rise in oil prices, in normal times - meaning that
> the GS CIF and other manipulation is not the main driver - the price
> of oil is reflecting the expansion of the economy, as it should.
> However, as one commentor suggested, and as I think too, a more appropriate
> look-see would be to examine the next 2/3 Qs that followed, and see
> what actually happens. Also, maybe 20% is the pinnacle as to where
> oil can rise in a 'growth tandem' with the economy, oil prices helping
> to support GDP - "to a point!"
>
> My question then, what is the effect on GDP when oil continues to
> rise above the inflection point of 20%.
>
> In my business, I see a rather large knee-jerk reaction to a 50 cent
> rise in gasoline prices, and have rolled up and down since mid-2005,
> following Bush's OIL COMPANY mergers, that have pushed us to a doulbling
> to tripling price tier ever since the mergers. Many business start
> to feel such pinches, gradually...however, 50 cents in gas does not
> necessarily translate, or stem from a 20% rise in oil....

Jul 14 10:20 AM

[The Inflect...:]

Thank you for the comment.

In the 1990 - 1991 recession there was a credit crunch which came before Iraq invaded Kuwait in August of 1990. It was not as severe as this current recession sure but they were pretty close together.

If you look at Q 3 1990, oil rose 130%, and that quarter GDP growth was 0. But if you look at the 1990 Q4 and 1991 Q1, the change in GDP went negative. (Please see below for these stats).

This brings me to the next point mentioned by many of the readers including you - the potential for a lagging effect that an oil spike has on GDP. I will take a look at this and come up with the results here or in another article.

Qtr Oil % Ch GDP %

1990q3 131.85 0.0
1990q4 -27.95 -3.0
1991q1 -31.07 -2.0


Thanks again for the comment.

Tom

On Jul 10 03:15 PM change is the only constant wrote:

> The overall analysis seems sound. However, if oil's price is both
> the source of growth and the reason for recession.... oil's price
> lead's and lags will feedback into the GDP numbers. Speculators causing
> a short term shock to costs may have a larger affect than can be
> quantified - it will depend on individual companies' ability to manage
> balance sheets with access to credit. Thinly capitalized companies
> will die with a severe shock. Adaquately capitalized finished goods
> producers at either full capacity, undercapacity and dealing with
> a credit contraction will all be affected by oil costs. The true
> effect oil has on GDP, and GDP on oil, gets into Black swan territory
> if you analyze the last 18 months like the previous decade. Never
> before has a oil shock been preceded by or followed by a credit crunch.
> Including a black swan event into your data stream about oil and
> GDP, that arguably includes a feedback bias, may leave you with only
> garbage.

Jul 14 10:49 AM

[The Inflect...:]

After I wrote the article about what if any link there was between an oil price move of 20% or more and a change in GDP, some wanted to know if an oil price spike could have a lagging effect on GDP. These comments made sense to me and I thought the same thing as I went through the initial thought process. So using only spikes up in price of 20% or more along with an analysis of GDP for the two quarters that followed the spike I have come up with the following:

Qtr Oil % Ch GDP %
1986q2 24.88 -
1986q3 - 3.9
1986q4 - 2.0

Qtr Oil % Ch GDP %
1986q4 21.97 -
1987q1 - 2.7
1987q2 - 4.5

Qtr Oil % Ch GDP %
1988q4 28.43 -
1989q1 - 4.1
1989q2 - 2.6

Qtr Oil % Ch GDP %
1990q3 131.85 -
1990q4 - -3.0
1991q1 - -2.0

Qtr Oil % Ch GDP %
1994q2 31.06 -
1994q3 - 2.3
1994q4 - 4.8

Qtr Oil % Ch GDP %
1999q1 37.23 -
1999q2 - 3.4
1999q3 - 4.8

Qtr Oil % Ch GDP %
1999q3 26.95 -
1999q4 - 7.3
2000q1 - 1.0

Qtr Oil % Ch GDP %
2000q2 20.77 -
2000q3 - -0.5
2000q4 - 2.1

Qtr Oil % Ch GDP %
2002q1 31.31 -
2002q2 - 2.2
2002q3 - 2.4

Qtr Oil % Ch GDP %
2004q3 34.24 -
2004q4 - 2.5
2005q1 - 3.0

Qtr Oil % Ch GDP %
2005q1 27.56 -
2005q2 - 2.6
2005q3 - 3.8

Qtr Oil % Ch GDP %
2008q2 37.84 -
2008q3 - -0.5
2008q4 - -6.3


So again there were 12 periods in which oil spiked by 20% or more. I have recorded the two quarters of GDP change that followed which means there were 24 quarters to look at. Of the 24 quarters, 19 had positive growth and 5 had negative growth. I am not sure of the effect that a spike of 20% or more in oil prices has on GDP even as a lagging indicator. That said if you look at two of the 5 quarters that saw negative growth, they came in Q4 1990 and Q1 1991 which were preceded by a 132% spike in oil prices. So a super spike of 100% or more in 1 quarter’s oil price could have a 2 quarter lagging negative effect on GDP.

Thanks again to all of the commenters for all of the great thoughts, and keep them coming. I / We enjoy discussions with some great market thinkers of SA out there. Comments that challenge my /our theses are especially welcome, as there may be something important in the analysis that I (we) missed in the first place.

Tom Henderson - Strategist JBH Capital


On Jul 11 05:49 PM G. L. Turner wrote:

> In the initial rise in oil prices, in normal times - meaning that
> the GS CIF and other manipulation is not the main driver - the price
> of oil is reflecting the expansion of the economy, as it should.
> However, as one commentor suggested, and as I think too, a more appropriate
> look-see would be to examine the next 2/3 Qs that followed, and see
> what actually happens. Also, maybe 20% is the pinnacle as to where
> oil can rise in a 'growth tandem' with the economy, oil prices helping
> to support GDP - "to a point!"
>
> My question then, what is the effect on GDP when oil continues to
> rise above the inflection point of 20%.
>
> In my business, I see a rather large knee-jerk reaction to a 50 cent
> rise in gasoline prices, and have rolled up and down since mid-2005,
> following Bush's OIL COMPANY mergers, that have pushed us to a doulbling
> to tripling price tier ever since the mergers. Many business start
> to feel such pinches, gradually...however, 50 cents in gas does not
> necessarily translate, or stem from a 20% rise in oil....

Jul 16 12:59 PM

[Stone Fox C...:]

Good info. Oil is definitely a coincident indicator. Oil prices also don't have the overall impact on the economy and inflation that most think. It hurts, but wages are by far the most important.

Jul 16 05:48 PM

[The Inflect...:]

Thanks for the comment. The numbers on the table did not translate well from my spread sheet to the SA site. My apologies to all the readers.

I will re-post the spread sheet.



On Jul 16 05:48 PM Stone Fox Capital wrote:

> Good info. Oil is definitely a coincident indicator. Oil prices also
> don't have the overall impact on the economy and inflation that most
> think. It hurts, but wages are by far the most important.

Jul 17 09:56 AM

[The Inflect...:]

After I wrote the article about what if any link there was between an oil price move of 20% or more and a change in GDP, some wanted to know if an oil price spike could have a lagging effect on GDP. These comments made sense to me and I thought the same thing as I went through the initial thought process. So using only spikes up in price of 20% or more along with an analysis of GDP for the two quarters that followed the spike I have come up with the following:


Qtr Oil % Ch GDP %
1986q2 24.88 -
1986q3 - 3.9
1986q4 - 2.0

Qtr Oil % Ch GDP %
1986q4 21.97 -
1987q1 - 2.7
1987q2 - 4.5

Qtr Oil % Ch GDP %
1988q4 28.43 -
1989q1 - 4.1
1989q2 - 2.6

Qtr Oil % Ch GDP %
1990q3 131.85 -
1990q4 - -3.0
1991q1 - -2.0

Qtr Oil % Ch GDP %
1994q2 31.06 -
1994q3 - 2.3
1994q4 - 4.8

Qtr Oil % Ch GDP %
1999q1 37.23 -
1999q2 - 3.4
1999q3 - 4.8

Qtr Oil % Ch GDP %
1999q3 26.95 -
1999q4 - 7.3
2000q1 - 1.0

Qtr Oil % Ch GDP %
2000q2 20.77 -
2000q3 - -0.5
2000q4 - 2.1

Qtr Oil % Ch GDP %
2002q1 31.31 -
2002q2 - 2.2
2002q3 - 2.4

Qtr Oil % Ch GDP %
2004q3 34.24 -
2004q4 - 2.5
2005q1 - 3.0

Qtr Oil % Ch GDP %
2005q1 27.56 -
2005q2 - 2.6
2005q3 - 3.8

Qtr Oil % Ch GDP %
2008q2 37.84 -
2008q3 - -0.5
2008q4 - -6.3


So again there were 12 periods in which oil spiked by 20% or more. I have recorded the two quarters of GDP change that followed which means there were 24 quarters to look at. Of the 24 quarters, 19 had positive growth and 5 had negative growth. I am not sure of the effect that a spike of 20% or more in oil prices has on GDP even as a lagging indicator. That said if you look at two of the 5 quarters that saw negative growth, they came in Q4 1990 and Q1 1991 which were preceded by a 132% spike in oil prices. So a super spike of 100% or more in 1 quarter’s oil price could have a 2 quarter lagging negative effect on GDP.


On Jul 11 05:49 PM G. L. Turner wrote:

> In the initial rise in oil prices, in normal times - meaning that
> the GS CIF and other manipulation is not the main driver - the price
> of oil is reflecting the expansion of the economy, as it should.
> However, as one commentor suggested, and as I think too, a more appropriate
> look-see would be to examine the next 2/3 Qs that followed, and see
> what actually happens. Also, maybe 20% is the pinnacle as to where
> oil can rise in a 'growth tandem' with the economy, oil prices helping
> to support GDP - "to a point!"
>
> My question then, what is the effect on GDP when oil continues to
> rise above the inflection point of 20%.
>
> In my business, I see a rather large knee-jerk reaction to a 50 cent
> rise in gasoline prices, and have rolled up and down since mid-2005,
> following Bush's OIL COMPANY mergers, that have pushed us to a doulbling
> to tripling price tier ever since the mergers. Many business start
> to feel such pinches, gradually...however, 50 cents in gas does not
> necessarily translate, or stem from a 20% rise in oil....

Jul 17 09:59 AM

[The Inflect...:]

I transferred the information again to make it easier to read. Here is the change in GDP for the 2 quarters following a 20% or more rise in oil.

Q2 1986 saw a 24.88% rise in oil, in Q3 1986 and Q4 1986 GDP increased by 3.9% and 2.0%.

Q4 1986 saw a 21.97% rise in oil, in Q1 1987 and Q2 1987 GDP increased by 2.7% and 4.5%.

Q4 1988 saw a 28.43% rise in oil, in Q1 1989 and Q2 1989 GDP increased by 4.1% and 2.6%.

Q3 1990 saw a 132% rise in oil, in Q4 1990 and Q1 1991 GDP decreased by 3.0% and 2.0%.

Q2 1994 saw a 31% rise in oil, in Q3 1994 and Q4 1994, GDP decreased by 2.3% and 4.8%.

Q 1 1999 saw a 37% rise in oil, in Q2 1999 and Q3 1999 GDP increased by 3.4% and 4.8%.

Q3 1999 saw a 27% rise in oil, in Q4 1999 and Q1 2000 GDP increased by 7.3% and 1.0%.

Q2 2000 saw a 21% rise in oil, in Q3 2000 GDP decreased by 0.50% and in Q4 2000 increased by 2.1%.

Q1 2002 saw a 31% rise in oil, in Q2 2002 and Q3 2002 GDP increased by 2.2% and 2.4%.

Q3 2004 saw a 34% rise in oil, in Q4 2004 and Q1 2005 GDP increased by 2.5% and 3.0%.

Q1 2005 saw a 28% rise in oil, in Q2 2005 and Q3 2005 GDP increased by 2.6% and 3.8%.

Q2 2008 saw a 38% rise in oil, in Q3 2008 and Q4 2008 GDP decreased by 0.5% and 6.3%.

Jul 17 10:29 AM

[The Inflect...:]

Correction from 1994:

Q2 1994 saw a 31% rise in oil, in Q3 1994 and Q4 1994, GDP increased by 2.3% and 4.8%.

Jul 17 10:32 AM