Bryan Caplan  

Arbitrage Challenge

Class Discussion Starter... If I Thought Oil Prices Would ...

Suppose the following:

1. You know now that the price of oil will be flat for five years, then fall by 10% per year every year thereafter. Everyone else thinks the price will be flat forever.

2. The longest oil-futures contract goes out 3 years.

3. A hundred other people will suddenly "see the light" and start agreeing with you one year from now.

Do you see any arbitrage opportunity in this market? If so, describe it in detail. If not, how can such valuable knowledge be so useless?

Comments and Sharing

COMMENTS (22 to date)
Chris writes:

Buy stock in a company that uses lots of oil, and short stock of a oil driller. Not arbitrage, which I think points to the answer to the second question - to get a risk-free profit, you need financial markets that allow you to isolate the risk that you want to bet on.

Horatio writes:

Wait 4.999 years and then start buying futures contracts. The price of those contracts shouldn't change over the next five years and you have better things to do with your money in the meantime.

I think Horatio missed Bryan's assumption #3. Competition with the other 100 will prevent you from getting a deal 4.99 years hence.

Maniakes writes:

Short oil futures whenever they exceed the spot price by enough to compensate for expected volitility and the cost of capital (based on your knowledge that the market's expectation of increasing prices is incorrect).

Sell 5+ year oil futures off of the publicly traded markets. Off the top of my head, you could do this by starting up a fuel supply company and offering long-term supply contracts with locked-in prices. The hundred people who see the light next year can invest in your fuel supply company to help you expand.

Carl Shulman writes:

How much capital do the 100 have available to invest?

JPC writes:

1. Ill-posed question, you can't know what everyone else really thinks past 3 years without a futures market.

2. Can be solved with private/OTC markets.

Eric writes:

It may be parsing the term 'futures' too much, but you could surely get an investment bank to write you a custom forward, option, or swap contract for a longer time period.

Futures are basically just standardized, exchange-traded forward contracts--and for them to exist the exchange would need a lot of liquidity to make them worthwhile to offer. A bank just needs to find a counterparty (and 'everyone else' thinks prices will stay flat, so that shouldn't be too hard, especially if structured as a swaption or an option on a forward, since they'd gladly take the free insurance premium--by writing a put--on a downside in prices they don't foresee) or find another way to hedge it synthetically.

The forward/swap market is huge, basically unregulated, and designed exactly for these type of unique/non-exchange-traded investments.

Stephen W. Stanton writes:

It would be impossible to arbitrage with only futures contracts daed 3 years or less.

The big banks make money in commodities by changing the rules of the game. For example, they buy physical oil, storage facilities, and tankers. They buy long-term supply contracts from crude suppliers. They sell duration-matched supply contracts to refineries.

They buy long-term supply contracts of diesel and gasoline from utilities. They sell long-term supply contracts to airlines and electric utilities. They then buy long-term electricity contracts from these utilities. They then sell long-term supply contracts to electricity users, such as aluminum smelters.

How is this relevant to your question? Well, you make money by changing the rules of the game. Find opaque and illiquid segments and make the market.

e.g., You could broker deals to sell oil storage facilities with a contract that gives the seller an amount based on the future price of oil vs. a fixed price. Structure your fees to increase when oil goes down.

In a nutshell, arbitrage is a tough, tough way to make money in the messy world of physical commodities. These are not paper assets that respond perfectly to models. They are messy, with qualitative differences between batches, with barrels going missing between point of shipment and the receiving terminal.

Some pure traders make money (lots of it). But the real money is made by the folks that rewrite the rules.

KF writes:

Because there are limits to the number of contracts that people can hold, 100 people initiating contracts in one year will not move the market significant enough, even with maximum leverage, to matter.

Therefore the question becomes, what does the other group of people think the price of oil will do over the next 5 years and then so on. If no one else knows what you know, then the price of oil shouldn't decrease, there is no arbitrage. If everyone knows what you know, then there is no arbitrage. If there is some function so that an increasing percentage of people that will come to this knowledge over the next 3-5 years, then they will begin to sell the long dated futures and buy the early dated futures. Therefore, your arbitrage (though not risk-free) today is to sell the long dated futures and buy the short-dated futures and keep rolling them over until the market has sufficiently marked down the futures relative to the cost of carry.

Scott Anderson writes:

I would buy a commodity whose price doesn't vary with oil but uses a lot of oil as input. Transportation comes to mind, maybe airline stocks. Plastic or plastic products like Rubber Maid also comes to mind.

Jim Glass writes:

Go to a prominent "peak oiler" and make a big-dollar public personal bet, as per Simon-Ehrlich. Go to as many as you can. See if you can get odds!

If you are serious, I think you'd be able to arrange private wagers and contracts of various kinds.

If your information really is infallible, you don't have to worry about losing and can eat transaction costs that would deter such arrangements with only imperfect information.

You can even leverage up as many times over as you wish, to get a huge payoff on a small margin after borrowing/transaction costs -- just like Long Term Capital Mangement did, but with the difference that you can only win!

Justin Ross writes:

Write and sell contracts that stipulate that one party will buy (or sell) a futures contract on a certain date in the future. Kind of like a futures contract on futures contracts. I think that is probably cheating on the spirit of your project.

j writes:

With current high price of oil, many companies are investing in substitutes that will become impractible in the future when oil will be cheaper. These companies are probably doomed, so they can be shorted. Traditional oil companies and refiners are underpriced, so they are good investment opportunities.

Globalization is dying, so land in faraway places will become cheaper. When New Zealand real estate have become really cheap, I would start to buy.

I would publish a book and become a prophet, and start collecting fees six years from now, when people had realized my wisdom. I would become a strategic corporate planner. Secretly, I would attribute my phenomenal foreknowledge to a personal demon (like Casanova did) and would diversify into life style advice and sell astral charts.

The word arbitrage has really gotten stretched in common -- and academic -- usage. At one time it meant a financial maneuver which cost none of your own money, but made money with 100% certainty (or about the certainty that a major first world government bond would pay off.), so essentially no risk at all.

Today, you see the word arbitrage being used all the time for opportunities which have a positive expected payoff, but aren't riskless, and may involve some initial outlay by you. By this definition of arbitrage, you could buy a diversified portfolio of stocks not including firms that benefit substantially from higher oil prices, and including in greater weights than otherwise, companies that benefit from lower oil prices.

The other thing is, if these are 100 are average people, they will not have access to enough funds, at least at first, to influence oil options and futures markets greatly, so initially, five years out, you will make a lot of money shorting oil.

Nathan Smith writes:

I don't think you can successfully arbitrage this.

Work backwards from 5 years from now. At that time, there will be outstanding 3-year contracts to supply oil at locked-in prices at some price. What price? Since prices will fall, anyone committed to buying oil at the current market price for three years will lose money, while suppliers will profit. So presumably, the outstanding 3-year contracts at that time will be offering oil at a discount relative to the current market price.

When will the discounts begin to be offered? At least three years before, when the price drop begins to fall within the contract period. Will they be offered before that?

Suppose it's one month before the new discounted contracts will become available. Buyers can get a long-term contract now, at the going market price, or they can get it in one month, at a discount. Will they want to wait? I don't see why, because-- if they foresee the price drop by that time-- they can just sign a three-year contract now with locked-in prices, then start buying oil on the market after that contract expires, just in time for the price drop.

"If not, how can such valuable knowledge be so useless?" The information concerns a good, oil-five-years-from-now, for which, ex hypothesi, no market currently exists.

Horatio writes:


I did neglect the third assumption. In that case, you could sell oil company futures now.

Tyler Cowen writes:

Roll over a series of successive one-year short positions, always keeping yourself on the short side of the market. Your margin collateral allows you to earn the T-Bill rate and sooner or later a big price movement will pay off in your favor, during at least one of those contract cycles. In the meantime you're earning the going rate of return.

8 writes:

I would go long the U.S. dollar and short the Russian ruble.

Joe Teicher writes:

This is easy. Sell strangles!!! The crude futures options market is very large and the little options (especially the calls) are bid to the moon. You know where every future is going to settle so you can risklessly sell any out of the money option.

Climateer writes:

Re: Mr. Cowen @ 07:22 AM:
Make sure your credit lines are in place.
Metalgesellschaft didn't fail because the stacked hedge was incorrect (it was textbook) but rather, because the directors freaked at the size of the cash drawdowns.
The forensic accountants said it would have worked, albeit with cash infusions in the billions.

Chris Neely writes:

A few thoughts:

1) Short stocks whose prices covary positivel with oil prices, like oil companies (I guess) while going long in companies that have the reverse exposure, like utilities that heavily use nuclear power plants.

2) Short oil futures contracts. When market expectations start to recognize that spot prices will drop, the futures prices will also drop due to the usual arbitrage arguments. It won't happen immediately, of course, but it must happen as markets begin to expect the oil price drop.

3) The CAD and other currencies have exposure to oil price movements. (See the work of Robert LeFrance.) Take appropriate positions in these currencies to exploit the expected drop.

4) Go long (short) in automobile firms that specialize in making large (small) cars.

Sorry I wasn't more specific, but I think that these will do it.

Of course, these are all risk-arbitrage. They require some capital and some risk tolerance but they should work if the price of oil drops in 5 years.

Am I missing something?

HalFin writes:

I just wanted to mention that in fact, oil futures presently go out 8 years. You can contract for oil to be delivered in 2016 for about the same price as today. Might not be such a bad idea?

Comments for this entry have been closed
Return to top