Category Archives: commodities

When is an end-user not an end-user?

When it’s a derivatives dealer.

On Tuesday, February 15, 2011, the U.S. House of Representatives held a hearing that focused on the regulations to implement the end-user exemption to the new clearing requirements for the OTC derivatives market. The New York Times covered the hearings, and reported that “the law exempted end users like United Airlines and Shell Energy from the clearing requirement.”

Is that true? Certainly the law didn’t expressly name United Airlines or Shell Energy. It simply exempted end-users—the very non-financial companies organized to produce real goods and non-financial services that trade derivatives to reduce the financial risk arising from their commercial activities. These are the types of companies and the type of risk management that this blog is expressly targeted to.

But I was taken aback to see the New York Times identify Shell Energy as an “end user” that “the law exempted … from the clearing requirement.” Of course, the average man-on-the-street thinks of Shell as a company with gas stations, refineries, storage tanks and oil wells. It is. But that’s not the whole picture. The average Wall Street investment banker running an energy derivatives trading desk also thinks of Shell Energy as one of its biggest competitors, dealing derivatives and running a proprietary trading business. Reproduced below is the awards list for the top 10 derivatives dealers in energy in 2009 as ranked by the financial industry magazine Energy Risk. Shell is at seventh place, following Morgan Stanley, Barclays Capital, Deutsche Bank, and others.

Surely the banks on this list don’t count as “end users” and neither should Shell’s dealer business. Shell the oil producer, refiner, shipper and marketer is an end-user, and Congress expressly drafted an exemption for this type of commercial activity—under certain circumstances and with certain provisos yet to be fleshed out in detail. But, Shell the derivatives dealer and prop trader? Why should they be exempted? Surely not because they happen to be owned by the same parent company that owns the wells, refineries, storage tanks and gas stations.

Continue reading

Insuring against snow storms with futures?

NPR ran a story this week on how airlines and other business are using futures contracts on snowfall “as a type of insurance policy against the weather.”

Weather derivatives can be a useful instrument for hedging, and hedging is a kind of insurance policy. But it is useful to keep in mind how a futures contract is different from an actual insurance policy.

A futures contract is written on a very specific commodity price or other well specified index. The snowfall futures contract is based on measured snowfall at a given location–such as Chicago’s O’Hare Airport–within a given month. It has a payoff that is very mechanically tied to that index. If snowfall impacts an airline’s business by cutting into revenues from flights, then buying a futures contract on the amount of snowfall may help reduce the total volatility of cashflow.

The catch is that the correlation between the loss of business and the amount of snowfall is imperfect. The futures contract doesn’t payout the amount of an airline’s losses. It’s payout is determined by the index, and the index isn’t always a good signal of the airline’s losses. The number of flights canceled is likely to be greater if the snow is concentrated in a couple of heavy storms than if it is spread out evenly throughout the month. But the futures contract payout will be the same in both cases. Similarly, two storms with comparable snowfall may have very different impacts on the complicated network of flights:

This is where insurance policies have an advantage. When I have an automobile accident, my insurance company covers the damage, whatever the damage–less my deductible. If snowfall causes roof damage to my house, my homeowner’s insurance covers the repairs. It doesn’t make the payout contingent upon total snowfall in the month, regardless of whether I have any damage or not. The payout is contingent on the scale of the damage. When hurricane Andrew struck Florida in 1992, causing enormous damage, it was the scale of the damage actually done that caused the large insurance payouts, not the measure of windspeed or quantity of rainfall. A hedging strategy using futures contracts doesn’t have this flexibility to make the payout contingent on the actual losses incurred. This is another example of basis risk.

Of course, in order for insurance contracts to offer this extra benefit of making the payout match the loss, they have to hire adjusters to investigate claims. Obviously there is room for a disagreement between me and my insurer over whether my auto or home claim is really covered by my contract and reimbursable, and what is the actual size of the loss. This distinctive feature of insurance–that the payout is directly tied to the amount of damage–is a costly feature. The transactions costs–in the form of adjusters, appraisals and disputes–are high. Futures contracts have low transactions costs by making the payout contingent on an easily measured index. But the tradeoff is that a payout linked to this index doesn’t quite match the losses.

The perils of hedging the price of jet fuel

Bloomberg had a news article earlier this week that highlights the practical problem of implementing a hedging strategy even in one of the largest commodity markets, jet fuel. Many airlines attempt to hedge some portion of the anticipated cost of jet fuel purchases over the coming year. In large part, they do this by purchasing futures contracts on either crude oil, the feedstock for producing jet fuel, or heating oil, another refined product that is slightly different from jet fuel. The prices of the three commodities are correlated, but only imperfectly. In the long run, if the price of crude oil goes up significantly, the price of the refined products must also go up. But in the short run, the relationship is very erratic. The price of crude and jet fuel can even sometimes move in opposite directions as they have recently. Here is a graph of the crack spread over the last number of years:

The spread graphed is the difference between the price of the refined products produced from a barrel of oil and the price of the barrel of oil. The graph shows the most widely quoted version of the crack spread, the 3:2:1 spread–i.e., 3 barrels of crude oil yields 2 barrels of gasoline and 1 barrel of heating oil. Clearly this spread swings significantly through time.

This is a classic case of basis risk, which is just the volatility in the differential between two related commodities or between one commodity for delivery at different locations or different delivery dates. Basis risk undermines the effectiveness of a hedge built using a related commodity. Airlines that used crude oil futures to hedge their jet fuel prices are finding that the hedges are losing money at the very same time that jet fuel prices are going up. This is exacerbating the cash flow volatility that the hedge is designed to reduce.

What can be done about this? Continue reading

Hedging with vertical integration — oil company edition

The NYTimes today carries a Reuters Breakingviews column arguing for a break up of the vertically integrated oil majors. The author trots out a number of questionable arguments in favor of a break up. He then voices a counterargument that is directly relevant to the themes of this blog:

“Executives argue that lumping refiners and explorers together offers a natural hedge. In theory, when falling crude prices hurt production, the refining business gets relief from lower input costs.”

Some might claim this is just a straw man, but I hear this thinking often enough that I think it is worth pushing back against it.

Two points.

First, in its simplest form, this argument just has to be wrong. At a first approximation, combining production with refining in no way hedges production. The oil producer is a natural long. The refiner is neither long nor short. The refiner both buys crude oil and sells refined oil products. It captures the margin between the two. Adding up the exposure of a producer with the exposure of a refiner yields a company that still has the exposure of a producer. There is no natural hedge there.

What I mean by ‘at a first approximation’ is that we assume that the price of crude and the prices of refined products all move together, one-for-one. Of course, this assumption doesn’t hold. When crude prices move, margins on refining often move, too. But the relationship is complicated, and in no way can it be described as a natural hedge. At best, it is like any two distinct operations within a broadly defined industry which have exhibited some negative correlations in the past so that profits from the two divisions seem to hedge one another. Whether the correlation is really there and reliable in the future is a very difficult question to resolve.

Continue reading

Futures Prices: The Weapon of Choice in Acreage Battles

The Financial Times reports that farmers are looking to futures prices as they decide which crops to plant:

Keith Phillips, a Missouri farmer, plans ahead. In November he determined how to plant his family’s 7,000 acres this year: mostly corn and soyabeans, with the rest alfalfa and wheat.
But a question mark remains against several hundred acres as he awaits signals from Chicago’s futures markets that could swing them from beans to corn. …

Many farmers have invested in fertiliser and seed and begun selling crops forward on futures markets, making planting plans inflexible.

This is, of course, exactly how it’s supposed to be. It’s always fun to see how mundane economic decisions can be spiced up with a good slogan: the FT headline calls this the “battle for acreage.”

In an article last month, the FT says that Cargill traders call this the “fight for dirt.” In this battle, futures prices determine the winner: “The planting projections will change as forward prices oscillate.”

Contagion in Natural Gas?

Wednesday’s FT has a story about the shuffling of US gas acreage among companies. Two things are happening simultaneously. First, the value of undeveloped shale properties has risen as the opportunities for profitable development become ever clearer. Second, while oil prices have climbed, natural gas prices have collapsed. Companies want to shift production to fields heavy in liquids and temporarily cut back production in fields filled with dry gas.

If companies already have acreage with liquids, great. Alternatively, if the company is cash rich, then it can afford to buy a new field with liquids. But companies with neither are under pressure to sell their dry gas acreage in order to fund acquisition of fields with liquids. Many companies are in this position, so plenty of properties are up for sale at the same time. Is that driving prices for dry gas acreage down even further than the fundamentals justify?

It would be a nice question to test empirically. There is plenty of academic literature about both fire sales and contagion. This could be another case study.

How much does the fire sale discount add to the value of companies with the liquidity to take advantage of it? How much does it subtract from the value of companies caught short of cash?

The Weak Tie Between Natural Gas and Oil Prices

The last two years has seen a dramatic decoupling of natural gas and oil prices. After prices for both commodities collapsed at year-end 2008, the price of oil recovered, but the price of natural gas has remained low. The ratio of the natural gas price to the oil price is at the lowest point in decades.

Ironically, this has happened just as a string of academic papers had firmly established that the two price series move together–for example, here, here and here. The International Energy Agency’s 2009 World Energy Outlook reprinted a table from one of these papers showing how an increase in the price of oil would be mirrored over the subsequent 12 months by a matching increase in the price of natural gas.

My colleague, David Ramberg, and I have just released a study attempting to make sense of these apparently contradictory facts. The bottom line: while a tie between the two series can be firmly established in a statistical sense, two features of this relationship have not gotten enough emphasis. First, there is an enormous amount of extra volatility in natural gas prices, so over short horizons like one or two years, the statistical tie between the two price series is very often swamped by extra swings in the natural gas price. Second, the level of the tie between the two prices — the normal ratio — is unstable over time, so that at longer horizons it is difficult to pin down the expected relationship.

%d bloggers like this: