« What I've been reading | Main | The fear of death »
My bet on oil prices
John Tierney is looking to bet that oil prices will fall. He can find some willing opponents here, and they will offer him the best odds available. I've been urging Alex to short real estate investment trusts; if he has already done so I fear for his ruin. Brad DeLong discusses Google. Jane Galt warns against such bets, on the grounds that timing is everything and no one knows when the bubble will burst.
It remains an open question why markets don't sell long-term bets for those people who have "price knowledge" but not "timing knowledge." The longest-term NYMEX crude oil futures sell for 84 months ahead; admittedly a forward contract can stretch longer. But why can't you make a 30-year bet on organized markets? I see a few hypotheses:
1. There are few if any people who have real price knowledge but not timing knowledge.
2. The disagreement in the market is about timing, so shorter-term contracts attract the attention and volume. You might disagree about whether something will happen this month or next, but you can't argue about whether it will happen ten or eleven years from now.
3. Long-term contracts make economic sense and someday we will have them. Right now we are out of equilibrium. We await tolerant regulators and heroic entrepreneurs.
4. You can replicate long-term contracts by trading short-term contracts in successive fashion. The informed traders have enough liquidity and borrowing power to make this work. (But hey, if that is so easy, why not have even fewer oil futures contracts?)
5. Liquidity is scarce, and involves significant economies of concentration. Intermediaries limit the number of contracts, just as we have limited trading locales and trading hours. (Admittedly many of these limits have, for better or worse, broken down.) Remember when the French used to trade stocks just a few times a day?
Take your pick or add to the list. Unless you opt heavily for #3, the market is saying that a general knowledge of price -- without a knowledge of timing -- simply isn't worth that much.
My bet on oil prices will be restricted to buying another economy car, next time around. It is better to spend the money on books anyway, no?
Thanks to Tim Bartlett for the pointer.
Posted by Tyler Cowen on August 26, 2005 at 07:18 AM in Economics | Permalink
TrackBack
TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d8341c66b253ef00e550834cbb8833
Listed below are links to weblogs that reference My bet on oil prices:
» Long Term Contracts from The Stone City
Liquidity is often supplied by proprietary traders and hedge funds. However, this liquidity never serves
to extend the forward curve to longer maturities. The main reason is that hedge funds must show results
to their investors relatively frequentl... [Read More]
Tracked on Aug 26, 2005 10:26:42 AM
» Bets on Oil Prices from Zmetro.com
Tyler Cowen summarizes a number of bets on energy prices. Cowen is buying an economy car next time around. T. Boon Pickens recently said that we'd see $70/barrel before $50.... [Read More]
Tracked on Aug 26, 2005 10:16:06 PM
Comments
The oil market reminds me a bit of the gold market in the late 1970s. Amid
predictions of $1000 gold and a coming economic apocalypse, the yellow metal
topped out at about $850 in January 1980, then began a long decline.
World oil demand is growing at 3.5-4% annually, even with China's 7% growth.
Assuming it grows at 4% it would take 18 years to merely double. How does
that translate into $200 oil in 2010 for an average price without a large
fall in it supply?
With the Dec. '10 contract under $64, $200 oil seems like a real sucker's
bet.
In the meantime, you can read Thomas Gold on _The Deep Hot Biosphere_ and
Peter Huber and Mark Mills on _The Bottomless Well_.
Posted by: Bill Stepp at Aug 26, 2005 8:52:42 AM
With a finger in the air, I like a combination of 2 and 5, where there is not enough disagreement about the LT price but lack of knowledge of timing as you stressed but combined with network effect for volume/liquidity premia.
Also, I would eliminate 6. In the presence of collateral posting, you only need to evaluate credit risk until next posting date.
Posted by: JS at Aug 26, 2005 9:08:19 AM
Hypothesis (4) is, I fear, wrong; this was spectacularly demonstrated by Metallgesellschaft AG in 1994.
More generally, rolling short contracts only works when the cost of carry (convenience yield of commodities) is deterministic. This is a reasonable approximation for stocks, but completely off the mark for oil.
Posted by: sammler at Aug 26, 2005 9:35:34 AM
#2. There's a lot of time-preference to make up for.
Posted by: josh at Aug 26, 2005 9:56:43 AM
Could it be that the probability of short-term fluxuations near the maturity date of long term contracts increase the risk of such contracts beyond acceptable limits for most potential traders?
A "bundled" contract for a certain amount of oil once a week for a year ten years from now could have a much more certain long-term value than a contract for all of the oil on the same day.
Having said that, a ten-year contract on oil, of any form, is much more a bet on policy and politics than anything else. What are the odds that nine years from now Iraq and Iran have a policy interaction (war or merger) which affects prices? Egypt and Saudi Arabia? What are the odds that the US engages in an agressive policy of building nukes or of exploitation of domestic oil sources currently held off-limits? What are the odds that the US falls into a period of significant inflation? What are the odds of a global recession?
I think that the lack of long-term contracts on oil reflects a fundamental volatility which is both of high magnitude and low frequency.
Posted by: Nathan Zook at Aug 26, 2005 9:59:46 AM
I have posted a followup at http://stonecity.blogspot.com/2005/08/long-term-contracts.html -- your trackbacks do not appear to be working.
Posted by: sammler at Aug 26, 2005 10:28:29 AM
How about "most people really don't focus much on the long term?"
Anyway, I don't see how a long term contract can work. If you use leverage, it can't really be "long term" if fluctuations in price make premature margin calls likely. If you don't use leverage, only radical predictions can have an expected return better than that of the market.
Long term uncertainty about nominal variables is certainly important.
I think that much finance silliness goes away if one reformulates the relationship between wealth and utility as a relationship between log wealth and utility. Doing so certainly clarifies the problems with leveraged financial strategies.
Posted by: michael vassar at Aug 26, 2005 10:32:23 AM
Traders are an essential element in the success of a contract market, but not the fundamental reason the market exists.
I would add to your list that there is no way to link the risk of a producer to a term longer than presently exists. Another way of stating it is that the existing time span of markets imply the longest time span in which producers can realistically plan.
Isn’t it true that in the real world a trading market must have investors and/or producers as a foundation? Real trading markets don’t exist without producers and/or investors who seek to minimize risk. I’m ignoring the current popularity of game markets like election future, etc.
A viable forward market is dependent upon the needs of producers and investors. Traders provide liquidity for hedgers. Pork bellies don’t exist to promote trading. They exist to promote production. Investors use CBOT options to hedge current positions, to optimize income, etc. The speculation by traders generates volume and thus liquidity for the hedger.
Posted by: chipper at Aug 26, 2005 11:29:10 AM
"the yellow metal topped out at about $850 in January 1980, then began a long decline."
I bought our wedding rings in January 1980. We celebrated our 25th anniversary this past April with our 3 children (23, 20 and 18) and numerous friends. The rings were cheap.
Posted by: Robert Schwartz at Aug 26, 2005 11:48:14 AM
My view is that 7 years is about forever. Any contract longer than that will be dominated by interest rates and inflation which can be effectively hedged in the long term financial instruments markets.
Tierney's bet is only for 5 years, so it can be replecated in the futures market, although not at the sucker's terms that he got. Nonetheless, it is important that very few peak oil mouths are putting up money to back up their hysteria.
Posted by: Robert Schwartz at Aug 26, 2005 12:03:06 PM
Mr Tierney got the odds much better than anything he would find on the futures/options market, by my rough estimates about 100 times better. Some rudimentary hedging would make his bet the best investment in his life.
Price knowledge without timing knowledge is, basically, useless. I can bet any amount at any odds that the price of any given commodity will be $1234 at some point in time in the future. I will never have to pay anything and there is some chance of winning - if the price gets there during my lifetime. A bet, anyone? You can pick the commodity, the odds, and the level as long as your bet is at least $100 ;)
30year futures contracts do not solve the problem: if am not sure about timing, can I be sure that the price will not climb to the desired level in 20 years time and then slide back?
Infinite maturity american-style options are appearing on the market, but their price is too sensitive to the carry cost and volatility.
So my guess is that long-term markets do not exist for a number of reasons mentioned in other comments, but even if they would exist they would not solve the problem you are discussing.
Posted by: mic at Aug 26, 2005 12:48:12 PM
As a wise banker once said, Oil prices will fluctuate.
What is more interesting is in what currency will these fluctuations be recorded.
http://www.informationclearinghouse.info/article9698.htm
Which bourse using which currency will dominate the trading and thus dominate the macro economics?
And just as a thought experiment, imagine that the bourse allowed not only euros but also yuan, and yen as acceptable currencies. Oil could become the real money that gold currently is. Oil discipline on national currencies.
Posted by: CK at Aug 26, 2005 1:11:00 PM
What is really odd is that the futures market prices range between $66 and $69/barrel, and yet a recent Reuters poll of analysts yielded and an average forecast of only a shade over $50/barrel. See here:
http://the-econoclast.blogspot.com/2005_08_21_the-econoclast_archive.html#112508617561303808
Posted by: The Eclectic Econoclast at Aug 26, 2005 3:58:41 PM
Tyler,
I think it boils down to supply and demand. If more people wanted the long term contracts they would create markets for them. Pratically speaking though, speculating 7 or 30 years out ignores many other fundamental things (alternative fuels) that could potentially make the bet a big loser. I came up with the following oil option prices assuming a current price of 66 1/8, a strike of 66, risk free rate of 3.429%, settle date of 8/26/05, expirations of 9/15/05, 9/15/12 and 9/15/35, call volatility of 50.097 and put volatility of 24.887:
Expiration Call Put
9/15/05 $3.15 $1.47
9/15/12 $27.14 $14.06
9/15/35 $33.98 $18.91
Given this risk/reward, no one in their right mind would pay these kind of premiums for the time values used, hence no market.
Posted by: Patinator at Aug 26, 2005 4:17:12 PM
Keynes already has your answer: in the long run, we're all dead.
Review your black-scholes: the value of an American option is unbounded in time. Put differently, the random walk underlying the derivation of the pricing formula means any price will be traversed with probability approaching one as time goes to infinity. Since participants in security markets must work with finite quantities of capital, that one obersvation reveals tradable options will have a natural "maximum term."
But it's worse than that. The actual distributions underlying real securities are much heavier-tailed than the log-normal distribution used for the canonical black-scholes closed form solution. Also, options holders bear counter-party risk. Your investment bank or Options Clearing Corp maybe be solvent today, but twenty years hence? Both of these effects increase volatility, loosely speaking, and shorten still further the natural maximum term for a widely tradeable option.
That said, I will sell you any option you want, put or call, on any underlying thing you care to dream up, provided you meet two assumptions:
(1) It's a European Option. No early exercise allowed.
(2) The expiration date is after the year 2100
Posted by: tylerh at Aug 26, 2005 4:32:18 PM
While it does not directly answer the question, a frequently overlooked aspect of the energy trading markets is the lack of credit quality. Many large players, particularly hedge funds and independent power companies that buy oil/gas to hedge electricity production, make heavily leveraged bets and have questionable credit quality. It's very risky to have one of these organizations as your counterparty on a long-dated transaction.
Posted by: Rob at Aug 26, 2005 7:27:02 PM
1. Eclectic Econoclast, there is no reason why analysts consensus should match the market. They are analysts, not traders, after all.
2. Patinator, it is rather unusual to price calls and puts with the same strike at different volatilities. It is arbitrageable.
3. tylerh, the value of an American call option on a futures is upper bounded by the futures price. Hence, it is not unbounded when the futures price isn't.
Posted by: mic at Aug 26, 2005 9:55:43 PM
If an option is priced by volatility, time value, and interest rates, then an option priced out twenty years is going to be almost all time value. Unfortunately, the link below only values options up to 999 days out, but you get the idea.
http://www.volatilitytrading.net/
That being the case, there isn't much point in trading options that are that far dated, because we agree about 95% of the price (time value and presumably interest rates), and then we're just arguing about the volatility, which is a relatively small portion of the price of such a long dated option. Of course, I'm totally guessing. Ask a professional.
Posted by: bjr@yahoo.com at Aug 27, 2005 8:28:56 AM
The problem with such long term contracts is that price discovery becomes too difficult given the risk of the contract. Traders, clearly uncertain about economic conditions 30 years in the future, will make the bid/ask spread so wide on any long-term contract that the profitability of trading such contracts becomes nonexistent.
Patinator - using such dramatically different volatilities for calls and puts makes no sense from a trading standpoint. Given your prices, I could sell your 30 year call and buy the corresponding put, making me synthetically short a $66 crude contract for a credit of $15.07. I buy futures to hedge my position, giving me no risk to price fluxuations. The cost of carry on the futures contract over 30 years is fairly low, given that a professional futures trader can easily obtain 10% margin rates, as well as cross-margining the price movement risk against the options position, which means as long as I keep my option position on, I'll never suffer a margin call. By hedging interest rate risk with long term bond contracts, I look to make about $9 per contract on a considerably lower investment, perhaps as low as $3 total given institutional rates. That's a healthy 10% annualized investment with far less risk than the stock market.
Posted by: Jason at Aug 27, 2005 10:10:26 AM
Who would take the other side? You might want to bet that GOOG will crap out at some point in the next thirty years, but nobody is obliged to bet against you. Anyone who has the opposite view to you is able to make their positive bet by buying GOOG and they get a liquid instrument. So why would they want something that offered them the opportunity to make the same bet but without liquidity?
Posted by: dsquared at Aug 27, 2005 4:04:25 PM
mic & Jason - Point taken about the volatilities. I didn't input them, they were just input defaults for oil on my Bloomberg Option Valuation screen. Today, the call and put volatilites are just about the same and the arbitrageable mistakes are gone. I did think it odd when I saw it but am not an oil expert and trusted the historical volatilites from Bloomberg.
bjr - On the 30 year contracts in my calculations, time value was 99% of the price.
Posted by: Patinator at Aug 29, 2005 3:55:51 PM
There are several reasons there aren't 30 year markets:
1. Even clearing houses can become a credit risk over a 30 year time frame. The potentially vast array of prices could lead to mark-to-markets that would bankrupt exchanges (which act as the middleman in all exchange traded futures and options).
2. The time value of money (including potential inflation) makes such long term contracts as much a bet on inflation and interest rates as on the underlying commodity.
3. A point that was touched on above - no-one needs to hedge or trade for such long time periods. The market provides enough contracts of long enough time period to be useful for hedging and trading, and for those that might require longer time frames will be able to roll their hedges forward as happens every day on futures exchanges.
Posted by: Simon at Aug 31, 2005 9:50:08 AM
liqingchao 07年8月21日
google排名
google排名
wow gold
wow gold
powerleveling
powerleveling
wow gold
wow gold
powerleveling
powerleveling
wow powerleveling
wow powerleveling
wow power leveling
wow power leveling
wow power level
wow power level
world of warcraft powerleveling
world of warcraft powerleveling
world of warcraft power leveling
world of warcraft power leveling
china tour
china tour
hongkong hotel
hongkong hotel
beijing tour
beijing tour
翻译公司
翻译公司
上海翻译公司
上海翻译公司
北京翻译公司
北京翻译公司
北京搬家公司
北京搬家公司
鼓风机
风机
风机
货架
光盘刻录
光盘刻录
光盘制作
光盘制作
光盘印刷
光盘印刷
红外测温仪
红外测温仪
超声波测厚仪
超声波测厚仪
超声波探伤仪
超声波探伤仪
频闪仪
频闪仪
涂层测厚仪
涂层测厚仪
电火花检测仪
电火花检测仪
google排名
仓库
仓库
仓库出租
仓库出租
物流园区
物流园区
集团电话
集团电话
四环素牙
口腔常识
口腔常识
口腔医生
口腔医生
网站设计
网站设计
多媒体
监控
监控
监控系统
监控系统
门禁
门禁
门禁系统
门禁系统
搬家公司
搬家公司
条码打印机
条码打印机
牙周炎
牙周炎
Posted by: wslmwps at Aug 21, 2007 2:11:27 AM
吴尊
阿穆隆
林志玲
尚雯婕
大人物
王睿
Mac DVD Ripper
火狐浏览器
Firefox浏览器
Posted by: hoojk at Dec 2, 2007 9:28:29 PM
資金を増やそうとするのに不動産投資をするのが手っ取り早い。日本で不動産で東京 賃貸をさがすのはきわめて難しくシステム開発は日本の会社が良い。
Posted by: 不動産投資 at Jul 18, 2008 1:04:29 AM
