Hello, by Interfluidity: Err... is this thing on? Am I back? I think I'm back. ...
The whole oil thing seems so, like, last month, although I notice there was some kind of deadhead revival in SoCal a couple of days ago. Some quick, crude thoughts: the whole "fundamental" vs "speculative" debate is terribly miscast, as emphasized most recently by Jeff Frankel (via Mark Thoma), but also by Tyler Cowen, and me too.
What I liked best about the Thoma / Krugman model is that it gave us four lines to think about, two kinds of demanders (people who want to burn oil vs people who want to store it) and two kinds of suppliers (people who suck oil from the ground vs people who drain their tanks). An imbalance of speculation on futures (more longs than shorts) creates incentives for people with tanks to fill them, potentially shoving up one of the two demand lines (the one on the left-hand panel of the Thoma/Krugman graphs). But four lines is a lot of moving parts. I think the really interesting line is the right-panel supply line. Rather than "speculation" vs "fundamentals", I wonder whether discretionary oil producers are flat-out producing as much as they are able, given the infrastructure currently in place, and whether over the past few years they have held back on developing capacity, or whether they are in fact eager to pump but hitting "peak oil" limits.
Either story is consistent with James Hamilton's fundamentals, although one might call unenthusiastic production "speculative" in a certain sense. In the end, I think Paul Krugman wins the debate he started, if it was the left-panel demand line driving prices, the only piece futures-buyers can influence, we should observe storage in tanks. As both Robert Waldmann and Alea's jck (in a comment) point out, paper speculators only persuade oil producers to leave the stuff in the ground when they drive futures into something close to strict contango, because producers enjoy less of a convenience yield than people with tanks. Inventory should build in tanks before it builds underground, if increased stock demand is driving the story.
It's important to note that, just because futures buying / speculative storage probably did not drive the great oil price boom of early 2008, doesn't mean it could not affect prices. Imagine, in Mark Thoma's discussion, that rather than a parallel outward shift in demand, the slope of the stock demand curve flattens as well. The "flatness" of stock demand maps to speculators' conviction that prices will rise. If speculators are absolutely certain that (the present value of) future prices will be higher than the current price, then they would persistently buy as much as would be necessary to pull the flow market price to the expected future price. In Mark's scenario, speculators effectively choose a quantity they are willing to buy at above the spot clearing price, and prices revert once their appetite has been sated. But speculators might choose price rather than quantity. (Still, if they do, we should see inventory build.)
Of course, explaining the rise in oil prices is passé. Now it's all about explaining the fall. Is it demand destruction? an incipient long run? declining inventories? increased production? a speculative bubble going "pop"? I dunno. Do you? (Maybe it's those evil short-sellers.)
In the month-ago discussion, Arnold Kling was the first to point out the connection between option values and the convenience yield. ... The kind of option a convenience yield represents is fun to think about. It is an option whose underlying is fluctuating calendar spreads, rather than prices. There is a lovely symmetry, in that there is a positive convenience yield both on having the commodity available, and on not having the commodity (but having a place to efficiently store it). If that seems weird, recall how some-strike call and put options both have positive value, even though when one is in the money, the other cannot be. We can even derive a relationship between the expected value of these two convenience yields somewhat analogous to put/call parity.
This all seems very retro now, a month is a long time, in the blogosphere and in financial markets. If I can avoid the lights in the sky, perhaps I'll come up with something more exciting to write about soon. ...