May 18, 2005
A good economic forecaster needs to tell a story that makes sense. We do not need to be Brothers Grimm or even Mark Twain, but we must be believable and consistent with economic theory.
Of the two recent stories about oil prices, Goldman’s prediction of price spikes to more than $100 per barrel is possible, but Merrill’s over $50 for as far as the eye can see is not. However, my story is different from either of those.
The economic theory that needs to be addressed is price elasticity. This describes what percentage change in quantity will occur for any given percentage change in price for any good and service. Actually, the observed price is the result of interactions of both demand and supply conditions, so both demand and supply elasticities must be considered.
The story that makes $100 price spikes possible is similar to the following: Some supply shock could push up oil prices. These price increases, in turn, lead speculators to believe the demand for oil is increasing (which is wrong, but such beliefs certainly are possible). They, therefore, accumulate speculative holdings of inventory to profit from rising prices.
At the same time, consumers believe oil product prices are going through the roof. Therefore, they keep their heating oil tanks full and top off the gas tanks of all their cars. Until this process is completed, it appears as if a surge in demand for oil has occurred. Thus, prices surge.
However, prices cannot stay up. Once the tanks are full, demand tapers off. Furthermore, higher prices lead to economic changes. More miles may be driven with the more fuel efficient vehicles. Thermostats may be adjusted. The larger purchasing power left at the gasoline station (and transferred to suppliers who are mostly abroad) will not be spent elsewhere.
Those unpurchased goods accumulate on shelves. Production is cut. Jobs are lost. And the number of people needing to drive to work are reduced. Over time, more fuel efficient cars are purchased and people adjust their distance to work by changing jobs or homes. Thus, the spiking prices soon fall back.
Actually, the $100 story is not that believable. To get a supply disruption to double oil prices, supply elasticities indicate the shortfall from trend growth must be more than 15 percent in the short run, and substantially more over time. Nevertheless, a speculative inventory surge coupled with tank topping could create a temporary price spike.
The problem with Merrill’s forecast is that they fail to realize that quantities demanded tend to fall over time for any given price rise. That is why the oil price surge of 1973 was followed by years of downward drifting prices. The surge in 1981 was followed by the oil price plunge of 1986.
The last time I estimated the responsiveness of oil quantities demanded to price changes, I measured a short run responsiveness of 0.2 but a longer term change of 0.7. Thus, the highest prices occur near the first price peak, and then prices drift downward until trend changes in quantities demanded and supplied finally become consistent with what the long term price elasticities are generating.
In the past two years, oil prices have virtually doubled. To sustain that level, we probably need to increase quantities demanded because of economic growth by almost 30 percent worldwide or remove almost 25 percent from quantities supplied. Those who are waiting for refinery capacity to fail would need to see about 10 refineries shut down simultaneously.
Moreover, too much has been made about no refineries being built since 1978. In the first quarter, refineries were operating at 90 percent of capacity, down from 91.3 percent during the fall. As any refiner will explain, there are ways to increase production from existing refineries without seeking the environmental permissions to build new ones.
To be sure, we would have more productive refineries if we could build state of the art units; but environmental realities probably make such construction unlikely.
So why are we inventing stories that deny the responsiveness of supply and demand conditions in the oil markets to changes in prices? Could it be that hedge funds have over-committed their resources to oil and Wall Street is worried that brokerage firms will lose if hedge funds do not make good on their bets?
Whatever the reasons, let us get forecasting stories that recognize the elasticities of demand and supply in the petroleum markets. (I still believe oil prices per barrel will be in the low $40s before the end of summer).