This is slightly off-topic, but when I hear that the big banks are buying physical commodities and storage facilities themselves--Morgan Stanley is the largest owner of heating oil in Maine, or whatever--and then I read about Lehman losing a bunch of money on its Inland Empire real estate investment, and, of course, always in the back of my mind is the memory of those fun days at UBS watching management dive head-first into exotic options and credit derivatives and the asset-backed securities market, trying at all costs to get our league table rankings up...

Well. You get my drift.

"Conceptually, a hedger is a party trying to shed risk, usually accepting some cost to do so, while a speculator willingly takes on risk, hoping to profit."

That is correct. However, it assumes that both parties are price-takers. While such an assumption is largely correct, there have been a number of situations recently where a well respected firm (that presumably also has a trading position in oil) announces that it is forecasting the price of oil to be at $xxx by a certain period, and this then becomes a self-fulfilling prophecy.

Much of the money in long-only ETFs is neither speculation nor hedging; it is diversification. It is not a belief that the commodity going up, nor is it a belief that the commodity is less volatile than the currency. It is simply a belief that the commodity is uncorrelated with other asset classes.

Or would you characterize diversification as a form of hedging?

MK — It could be a bubble, in the sense that a sharp price move might shake people's conviction in the relative values of dollars and oil. That's not an easy thing to say. If it is a bubble, all the late chasers will get stuck with the bill. If the dollar is hopelessly worthless relative to oil, it's better late than never. Without more volatility to test people's convictions, it's really hard to know.

But the financials might be fully hedged. That is, if they are building storage capacity plus capacity to deliver, and if they are getting good spot deals, (which they very will might, because spot markets are messy and dealmaking matters) it is a heads I win tails you lose bet. They store the oil and sell it forward at a profit given current prices and storage/delivery/financing costs. When roll time comes around, they compare the profit of delivering immediately to the next roll price less total storage costs. Eventually they pull the trigger for a profit. At least that's how things are supposed to work.

Hedged shorts can become speculators when, at current prices, delivery is more profitable than rolling, but they choose to not deliver, and become an unhedged long hoping the futures price will appreciate. If they're right, by locking in a higher futures price they can do better than the foregone profit from the undelivered prior contract. If they're wrong, well, speculators who bet wrong get hurt. But, if the holders of oil are producers, there's a bias in the decision, to speculate or not to speculate when there's a contract I could profitably deliver. Not speculating, delivering the contract, satisfies real demand and limits price appreciation, while speculating into the next contract helps fan the flames. Since they are also selling in the spot market, or writing over-the counter contracts benchmarked to futures prices, oil producers have added incentive to speculate a bit, going briefly unhedged until prices appreciate to where I can lock in at least my foregone profit. If there is any "market manipulation" in oil, I think it is this sort of behavior by producers, failing to deliver when they would achieve greater risk-free profits from doing so than from rolling, in tacit understanding that other producers will do the same. The best way to combat such a tacit cartel would be to bring new players into the market interested in short-term transactional profits rather than maximizing the value of a large pool of production capacity. It'd be nice if financials took on that role, collecting arbitrage profits as large and as early as possible. It's hard to know what they would do. If they've secured sufficiently much access to cheap physical oil, their incentives might line up with producers. If the oil they store and sell forward represents most of their capacity, they might lean against the bubble, potentially popping it as cartel speculators get burned when futures prices fail to continue to rise...

anon — Does the above (disorganized, thinking aloud) help? It's worth pointing out that there is no perfect relationship between futures and spot prices given geographical dispersion of the physical spot product, and frictions (transport/storage/delivery capacity constraints as well as industry structure/incentives to arb as outlined above) that restrain arbitrage. In a frictionless world, futures prices and spot prices would converge exactly, and the relationship between the two would be set by costs of storage, projected "convenience yields" (opportunities to profit by lending the commodity prior to delivery), and the risk free interest rate. In a world with frictions, there is no one spot price, lots of different grades of fuel, different transport options and costs from and two different places, etc. Hopefully, the theoretical, frictionless pricing holds at least approximately. But "approximately" hides a lot of play for fortunes to be made or lost.

RTD — If firms are not price-takers, than there must be some kind of monopoly/market power that prevents competition from forcing price down to expected marginal opportunity cost. There certainly could be, as outlined above, if there is a game being played by producers that amounts to a tacit cartel, there would be monopoly power over price. Publicly announced price targets could help provide benchmarks around which to organize that kind of behavior. Of course "expected marginal opportunity cost" is kind of a slippery phrase, one way of saying what I just said is that, given scarcity rents, public price targets help to define expected marginal opportunity costs. (And expected marginal opportunity costs would have to include the "secondary" effects of transacting on the value of the rest of ones production capacity.)

Nemo — Though I should have made it clearer (I'd have had to have talked about the covariance as different assets as well as their volatility with respect to consumption bundles), I most certainly do consider mere diversification to be a form of hedging. Typically (though not always) purchasing diversifiers reduces ones expected return, but reduces portfolio variance relative to your benchmark. That is, mere diversifying trades risk for return, as is typical of hedging. (It's harder to characterize parties who expect an asset to perform better than the rest of their portfolio and covary little or negatively with other assets — they are expecting both to shed risk and enhance profit, so they are both speculators and hedgers by my lexicon, maybe depending somehow on motivation. But, absent unusual information, these opportunities should be pretty rare for most investors, if markets are not outrageously inefficient. (Markets might be outrageously inefficient, but most investors, particularly professional money managers diversifying into an asset class, probably form expectations as though they are reasonably efficient.)

FC — Thanks.

Benign — Throw enough dressing on them, and try not to notice if GW's face doesn't quite live up to the slang term "lettuce". (At least cake tastes good, ya gotta give that to Marie Antoinette.)

I'm running and not editing, so if I got something wrong in the stream of consciousness futures market conjecturing above, apologies in advance!

Mike:

No.

From Mauldin:

"The only reason to raise rates would be to protect the dollar from a serious collapse. I think it more likely the Treasury would intervene in the markets to prevent such a collapse. Dennis Gartman, at dinner Wednesday night, suggested that if the administration really wanted to get the market's attention, they could intervene in the currency markets and release oil from the Strategic Petroleum Reserve at the same time. While it would only be a temporary fix, it would make speculators nervous. However, they might consider such an experiment preferable to having the Fed raise rates during the middle of a slowdown/recession."

Boy is there a lot to talk about here.

anon — yeah, I think a speculator is anyone with a position unhedged, either by other positions or by future consumption or production plans. Financials who buy spot oil and sell it forward for a predictable profit are hedged arbitrageurs, until and unless they decide not to complete the arbitrage, and hold the oil for another round without immediately selling forward. Then they are "gap speculators", and I think they are key to oil prices right not. These are players who speculate, but hope to do so only for a brief period of time. They let their hedges expire without offloading physical oil, but instead wait until price momentum permits them to establish a new hedge that locks in greater profit. In any "greater fool" strategy, the art is to join the bubble, but be exposed to risk in it only briefly, to limit the probability that you are unhedged when it gives. That's the game I think those long physical oil are playing right now. (I want to hedge myself a bit, by adding the condition "if oil is a bubble". Dollar (or even Euro) prices don't mean very much right now, given the loosened coupling between currencies and goods & services. I want to look at the price of oil in terms of other commodities before opining definitively that it is a bubble.)

Whether or not speculators can affect spot prices without affecting inventories is a murky question. Krugman's clean supply/demand graph hides the complexity of spot markets. What is the spot price of oil? It depends on the grade you want, where you are, who you know, and how good at operations management you are. Even putting all these complexities aside, inventories are absolutely not measurable. A reasonable definition of inventory build-up in this context is the difference between the quantity of physical oil consumed in a period and the maximum achievable production capacity given existing infrastructure. We can estimate the former, but we have no idea about the latter, so we really cannot say whether inventory is building. If you haven't seen it, check out the impulse response graph by US producers to a negative oil shock by Francesco Lippi (via Mark Thoma). US suppliers diminish production following negative supply shocks (as opposed to positive demand shocks), indicating either producer speculation or tacit cartel price manipulation, depending on how you want to spin it. In either case, reduced production qualifies as inventory build in this context.

The overall picture is that all markets have to clear, so in theory you can look at spot in isolation (as Krugman did), and say here's supply, here's demand, the price is determined by these curves period, unless someone is building inventory. That's true. But building inventory doesn't show up (as it did in Krugman's sketch) as a third line in the graph. It shows up as a spike in demand (people buy oil to fill warehouses) or a decrease in supply (it's harder to get producers to sell their oil). We can't distinguish "real" from "speculative" shifts in supply and demand in real time. And we can't just look at tanks and tankers, because that's only one form of storage, and almost certainly the most minor one if in fact this is a speculative phenomenon.

benign — You'll have to flesh out a bit what your 1 period model looks like. It's hard to think in one period. On a given day, it's certainly true that price insensitive long interest in futures markets will drive up trading prices, as people need to be induced to take up unhedged positions on the short side. But those unhedged shorts, who were responding to long demand and probably are not interested in holding speculative short positions will fairly quickly try to lock in a profit by hedging, with some other financial position if available, or more slowly with physical exposure. Liquidity providers selling to long speculators are probably doing deals, making arrangements with producers for call options on oil to hedge their exposure, struggling to make operational arrangements so they could take and deliver if necessary. Producers writing calls have to hedge by making sure they can deliver if the options are exercised, either by storage, or planning for conditional excess capacity (which is equivalent to storage). All of this takes time, so in the day of a price spike, it is fair to say that speculation has changed prices without changing inventories. You might argue that the period since last August is like a long day, prices have moved faster than than operational arrangements those moves compel. Could be. But still, the forces at play are fairly instantaneous. The minute an unhedged long bids up a price, at the margin somebody is more likely to withhold then to sell spot oil than they were before.

RTD — Beautiful: "It still seems to be a case of the invisible foot getting in the way of the invisible hand." But...

I think you need to distinguish between commodities. Small farmers are price takers, and are definitely shifting production from lower priced to higher priced crops. But energy producers may not be price takers. Individually and through tacit collusion, they may have market power.

And for producers whose resources are bound to a single fixed-supply commodity (e.g. Saudi Arabia rather than a farmer who can pick crops), the relevant question isn't the price in absolute terms, but today's price versus future prices, combined with a discount rate based on how productively financial assets from a sale can be stored. An oil producer rationally cuts back production today if she believes prices will be higher tomorrow than the future value in real terms of saved funds. (The whole Hotelling thing, right?). So, even if prices are rising, but you view financial discount rates as falling or even negative, you're better off not producing.

If it's collusion, we should break it. But if it's producer timing, that's the invisible hand doing what it should be, no invisible foot required. Until we persuade oil producers that they have something productive in real terms to do with their cash, they're better off slowing down. Since many oil producers are peg/politically constrained to investing great sums in USD financial assets, well, I wouldn't be producing breathlessly under the circumstances.

The oil world is becoming somewhat clearer to comprehend as a result of your writing, thanks. You've described two critical factors that I had not appreciated:

a) The ability of financial investors (e.g. pension funds) to control long positions in physical oil. This seems fundamental to the whole discussion. I had assumed that investor participation in commodities as an asset class (e.g. oil) was done primarily through the futures market (e.g. buying futures and rolling from there). If this were the case, it would be difficult to understand how their participation was affecting the spot price. This was the motivation behind my interest in how the futures price can affect the spot price.

b) The inability to track above ground inventories of physical oil. In this case, Krugman's analysis becomes misleading, as you describe.

Everything else flows from these factors, and it becomes obvious how investors along with producers can affect the spot price of oil. Your examples showing tactical choices with respect to futures market hedging and other strategies are then quite clear. Whether investors are classified as speculators or hedgers, or whether they hedge their long positions in the futures market, is less relevant than their potential to affect the spot price of oil through physical positions. Moreover, the distinction between producer and investor becomes almost as misleading as the distinction between hedger and investor, insofar as the potential effect of physical market access on the spot price of oil is concerned.

It is equally obvious how producers can affect the spot price if they store oil instead of selling it. As with financial investors, whether or not they take open unhedged (i.e. speculative) long positions or hedge their positions in the futures market with arbitrage profit, is less relevant than understanding their potential influence on the physical market and spot market price.

To clarify further, my underlying question was whether or not financial investors could affect the spot price of oil as a result of their presumed (erroneously) exclusive activity in the futures market. This is somewhat different than the (more real world) question you addressed, which was “Whether or not speculators can affect spot prices without affecting inventories”.

If you can't track inventories, then you have no way of knowing how producers and investors are affecting the supply of oil on the spot market through inventory hoarding. So Krugman's neat supply/ demand/ wedge graph becomes irrelevant. And if, as you say “the maximum achievable production capacity given existing infrastructure” is unknowable, making inventories unknowable, then obviously the contribution to the spot oil price from either producer or investor hoarding of physical oil is unknowable.

It seems from widely reported figures that trend growth in global supply has been decelerating, while natural demand from developing countries has been accelerating. Obviously the net effect must be upward price pressure. But trend growth has still been reasonably smooth, albeit flattening, as I recall. Inventory hoarding, whatever its true nature and effect, doesn't seem to be in evidence as some recently observable kink in the ongoing trend of global consumption. The hyperbolic price action over the past 6 months still seems wild and crazy relative to this fairly smooth trend line of reported consumption, which by definition should be net of true inventory hoarding.

"June 16 (Bloomberg) -- On Wall Street, where just about everyone has lost confidence in financial assets, Goldman Sachs Group Inc. and Morgan Stanley are making money the old-fashioned way: Buying and selling commodities.

Goldman and Morgan Stanley are expected by analysts to report the best second-quarter earnings of the world's biggest securities firms this week, having limited their losses from the collapsing credit market. They also lead Wall Street in commodities trading, where crude oil futures doubled in the past year and the price of products from gold to corn soared to record highs."

I see a political/geopolitical conundrum coming quickly this fall. Israel wants to take out Iran's reactors before Bush is gone. While domestic politics calls for an oil crash ~2 months before the election.

What type of deal can be put together to accomplish both of these objectives at the same time?

Keep an eye on aircraft carrier movements, Saudi production and GS oil futures.

Steven,

I am a long time fan.

You say in this post that "The "wealth-holding class" responsible for the misdirection of capital will do what it can to shift losses to dispersed and relatively powerless little guys."

I would love it if you could either point to concrete examples of this or engage in a theoretical discussion of what they can do to disperse their loses as described. It goes to the core of something I've been thinking about, but I have yet to come up with any real compelling evidence to back up such a statement.

More from Krugman on my favorite question - the effect of futures on spot:


June 23, 2008, 4:38 pm

Speculative nonsense, once again

OK, one more try.

First of all, I don't have a political dog in this fight. I'm happy to believe that crazy speculation distorts markets. And I do think it's likely that oil prices will come down, for a while, once consumers have a chance to respond more fully to high prices by changing their driving habits, switching to smaller cars, etc..

But the mysticism over how speculation is supposed to drive prices drives me crazy, professionally.

So here's my latest attempt to talk it through.

Imagine that Joe Shmoe and Harriet Who, neither of whom has any direct involvement in the production of oil, make a bet: Joe says oil is going to $150, Harriet says it won't. What direct effect does this have on the spot price of oil — the actual price people pay to have a barrel of black gunk delivered?

The answer, surely, is none. Who cares what bets people not involved in buying or selling the stuff make? And if there are 10 million Joe Shmoes, it still doesn't make any difference.

Well, a futures contract is a bet about the future price. It has no, zero, nada direct effect on the spot price. And that's true no matter how many Joe Shmoes there are, that is, no matter how big the positions are.

Any effect on the spot market has to be indirect: someone who actually has oil to sell decides to sell a futures contract to Joe Shmoe, and holds oil off the market so he can honor that contract when it comes due; this is worth doing if the futures price is sufficiently above the current price to more than make up for the storage and interest costs.

As I've tried to point out, there just isn't any evidence from the inventory data that this is happening.

And here's one more fact: by and large, futures prices over the period of the big price runup have been slightly below spot prices. The figure below shows monthly data from the EIA; as the spot price shot up, the futures price (that's contract 4, the furthest out) actually lagged a bit behind. In other words, there hasn't been any incentive to hoard.

As I've said, I don't have a political dog in this fight.

But the nonsense in this debate makes me want to shoot someone in the face.

Update: I see that Michael Masters, about whom I had some flattering things to say a few days ago, is now telling Congress that gasoline will go back to $2 a gallon if we crack down on speculators. He forgot to mention that cold fusion will solve all our energy problems any day now.