The relationship between the weather, seasons, temperature and the price of energy must be one of the most misunderstood issues in what I'll charitably dub "The economics-media" (when journalists attempt to report on economic issues). How often have you heard a sentence like this: "The price of natural gas has risen x-fold this year, and we haven't even hit the winter home-heating season!"? Or how about this classic: "Gasoline prices are expected to stay high until labor day, the traditional end of the summer-driving season."?
I'm prompted to write this after reading stuff like this:
Heating oil and gasoline rose marginally to $1.6831 a gallon and $1.4605 a gallon, respectively. Natural gas rose 4.7 cents to $11.610 per 1,000 cubic feet.
Still, Phil Flynn of Alaron Trading Corp. suggested prices could soon rise, once temperatures drop.
"The debate is how cold will it get and for how long," Flynn said in a research note.
Forecasters are predicting much lower temperatures in the U.S. Northeast -- and thus higher demand for heating oil -- in the coming week.
Now, I'm going to give Phil Flynn the benefit of the doubt, and assume his research report was misunderstood by the journalist writing the story, but the idea that prices could shoot up as soon as the temperature drops next week is just silliness. Natural gas prices don't shoot up when it gets cold for the same reason that shares of H&R Block don't spike up right around tax-season--intertemporal arbitrage.
James Hamilton explains this well in the video I've been linking to, but it can be explained rather easily without watching a 35 minute lecture. Suppose you are the owner of a natural gas well, for example, and the stuff is trading at $11.6/1000 cubic feet of the stuff. But you think that in a few months it will be trading $15/1000 cubic feet, because you think the economy is going to accelerate faster than people think, and demand will be higher than expected. You're not going to bother drilling your stuff today, instead, you'll take a few months vacation, and return when you can sell it for 50% more. However, the act of all these people refusing to sell today will shift the supply/demand equation and thus cause an immediate rise in prices, to something closer to the $15/1000 that you want to get. This is intertemporal arbitrage--meaningful differences between price and price-expectations get smoothed out by market-forces. Everyone already knows that winter will get cold and demand for natural gas will increase. Today's price reflects this anticipation.
It works in reverse too. Take the "summer driving season" example we gave earlier. Since all of the gas companies know that gasoline demand will fall off after labor day they have an incentive to sell a lot of it during the summer so that they're not left with excess inventory during the winter. This creates downward pressure on prices, and thus prices don't collapse come the fall.
The best demonstration of the media's foolishness is a Simpsons episode in which Homer attempts to corner the pumpkin-market in the months before Halloween. Through a broker, Homer has the ingenious scheme of buying pumpkin-futures in the summer, well before Halloween. As October 31st approaches the value of his pumpkins skyrockets and his broker urges him to sell. Of course, in Homeric fashion, he confidently predicts the price to continue rising even after Halloween only to lose all his money when the prices crash.
If you listened to the media, we could all be Homers (sans the losses), merely by buying heating oil in the summer, and selling it as the temperature drops. Sadly, that reality exists only in cartoons and in the brains of cable-news talking heads.
What you are saying is true, yet the market reacts to the weekly supply numbers. Personally I believe we are in a multi year move in oil etc. because of a lack of investment from 1990 to 2000. Any comments? Thanks for the great site.
Posted by: Harry | November 16, 2005 at 11:33 AM
Well, to some degree (heh), the current temperature is important.
If we had severe weather early - major snowstorms in the NE US right now, for example, we could reasonably expect more heating degree days and therefore more energy consumption for the whole season, couldn't we?
The longer current mild weather lasts, the less cumulative consumption is expected by the market for the whole cycle. Not very precise, but kinda makes sense...
Posted by: Igor | November 16, 2005 at 04:19 PM
To both of you, yes, prices can change as new information comes to light. That's why the price does change on a day-to-day basis. And it's true, if winter comes two weeks late, then that's two-weeks worth of extra natural gas inventory which will put downward pressure on prices.
My complaint is that people have a very narrow-minded view "prices go up when it gets cold". That's actually not narrow-minded, just plaing wrong.
Posted by: The Stalwart | November 16, 2005 at 04:32 PM