I've often wondered about quantifying the impact of crude price shocks on the economy. I've held a belief that crude oil price shocks are the biggest reason we've had economic downturns in the recent past. So, I decided to take a look at some basic data, keep the analysis simple and transparent, and see what insights we can gain.
Crude Price and GNP Data
I am using quarter to quarter changes in real GNP (US Department of Commerce) and quarterly average crude prices (from the Energy Information Administration). Quarterly data takes out some of the monthly noise, and economic variables take time to have an effect. It's also a lot easier to interpret and analyze.
As some of my former colleagues know, I don't take a lot of stock in anectodal interpretations of why prices are or aren't moving a certain way. It's not hard to find anectodes after variables change, but I've yet to find anyone who can consistently and accurately look at the information in the market right now, and tell me what impact that will have on the price at the end of the day. The price to me simply reflects all available information out there perfectly (Chicago school style).
I am simply going to look at two variables: the price of crude, and economic output, and study their relationships, during recessionary periods.
Analyzing the shocks and after effects
After a casual look at crude and GNP, it is amazing that every single quarterly decline in GNP since at least 1974 is associated with an increase in crude prices. Maybe not cause and effect, but just saying, it is pretty clear that the health of our economy is super dependent on oil/energy prices.
The chart above and visible in a window here tracks three oil price shocks that occured since1979. In the chart, each shock event is anchored to the point prior to the increase in oil prices. Crude prices shown are differences from the quarter to the anchor point. The length of the line represents the total time of the episode - from the first crude price increase, until full economic recovery back to the level before the shock. Dashed lines represent quarters of economic decline, while solid lines represent times of economic growth.
Shock 1: 2Q90 - 3Q91
A relatively small shock - prices rose $20/B for two consecutive quarters, before falling back to historical levels. The economic impact was immediate, with GNP declining for two quarters (-1.2% - not shown in the graph), and then growth resumed, with full recovery in another three quarters, by 3Q91.
Shock 2: 1Q79 - 1Q83
A much larger shock, with prices rising at their peak to $50/B above the historic levels, with most of the rise occuring over a year. It took about a year and a half before economic output began to fall. At that point, there were some ups and downs, with some increases and decreases in GNP. The total economic decline was -2.3% and it took the economy 9 quarters to return to it's initial level prior to the decline.
Shock 3: 1Q07 - 1Q11
The most recent shock, which is still familiar to us. Prices rose over $60/B over the course of a year and a half. Prices dramatically fell, and then began another rise, albeit at levels about half the peak increase. The economy began to decline a year after the increase in crude prices, with one quarter of slight growth, before a continual decline in ouput lasting a year. The total economic decline was - 3.4%, and it took 2 years for the economy to get back to its peak level. With crude about $95/B, it continues to be about $35/B above the anchor point in the chart.
Pulling it together - key findings
So what are some overall conclusions? Here are some of my observations and intepretations, based on commonality across these three shocks. Reference the table here comparing stats across the three shocks.
Can we quantify what an oil shock is? The smallest of the shocks studied was $22/B in total over two quarters (1979). The biggest shock was $65/B over 5 quarters (2007). So we should be wary if crude rises $11-13/B per quarter or so for 2-3 quarters. That would be enough to potentialy cause a recession. Also, the absolute crude price needs to be relatively close to prior highs.
Every $10/B increase in crude, persisting for a quarter, hurts GNP by .5%. So a price shock of $50/B means we can expect economic decline of about 2.5%.
It takes about a half to three quarters of a year before the economy pulls into a recession after a price shock (note that for the three shocks, growth prior to the shocks was in the 2-3% range).
The economy can get over any price shock, eventually. The rate of recovery, following economic decline is 0.4% - 0.5% per quarter. Unfortunately, it's a slow road to recovery, but we eventually are able to conserve or use less energy, even if higher prices persist. We use almost half as much energy per GDP output today then we did in 1980, and that trend will continue.
At current crude prices of about $90-100 a barrel, we are going to continue growth, albeit slower than we might like. Based on this analysis, if crude prices rise to $120 and hold for at least two quarters, we are at risk of having a decline in economic output.
Compared with other oil shocks, crude prices remain relatively high currently. Given that growth after economic recovery (2.1%) was lower then growth before the price shock (2.8%), crude prices may well be dampening growth rates (at least temporarily).