The Future of Oil Prices

One of the topics which I have been following for the past couple of years is “Peak Oil”, the theory that global oil production will soon peak, imposing vast changes on the social and political world order. I find it a great test case for the issues we have been discussing here. The issue brings widespread disagreement, and it’s one where knowing the truth is of great importance even for the average person.

A good starting point to get a handle on the situation is to look at what commodities traders call a “futures strip” for oil prices. Here’s a chart I created showing the prices of oil futures contracts from January 2007 through December 2012 as of the close of trading yesterday:

Oil_20061222


The obvious message of this chart in terms of future oil prices is that they are expected to be relatively stable. The whole next six years stay within the range of $60-$70 per barrel. This is unlikely to be consistent with a significant oil shortage in that time frame. The markets seem to be telling us that Peak Oil is not a near-term concern.

GD Star Rating
loading...
Tagged as:
Trackback URL:
  • http://profile.typekey.com/robinhanson/ Robin Hanson

    Presumably these are expected values; it would be nice to also see estimates of variance.

  • Kaj Sotala

    At this risk of making a fool of myself – for the benefit of us non-expert readers, how were the values on this chart calculated?

  • http://profile.typekey.com/nicholasshackel/ Nicholas Shackel

    Why exactly is the peak oil issue anything other than another piece of alarmist scaremongering? I expect there was a time at which horse manure production peaked, but its peaking was of no significance. I certainly don’t see any reason to think it would bring vast social and political changes, unless politicians start imposing price controls.

  • http://profile.typekey.com/halfinney/ Hal Finney

    Kaj – These are the actual reported closing prices of the various oil contracts yesterday. Contracts exist for future oil purchases for every month out through 2012. Now, I think that there must be some massaging of the data before it gets reported, because of the suspicious smoothness of the curve. Nevertheless I assume that this is a good approximation to the market clearing price for oil contracts for delivery in those months.

    Robin – Good point about volatility. James Hamilton of the http://www.econbrowser.com blog did the calculation a few months ago:

    http://www.econbrowser.com/archives/2006/02/oil_at_1530_a_b.html

    He determined that for a 95% confidence interval on oil prices at the end of 2010, the range is $14 – $250! So it would be more accurate to say that the markets are not expecting Peak Oil, but they are not ruling it out either (nor are they apparently ruling out a glut).

  • Perry E. Metzger

    I’m sorry to break this, Hal, but you’re off base here. All that is saying is what the market thinks interest rates are like right now. The price curve for all futures looks like that, always. (Well, not *quite* always, but often enough that this is the universal expectation.)

    The reason for this is fairly straightforward. Futures prices always reflect current prices modulo expectations on the price of money because you can always arbitrage a contract in time if you have cash available, and you can always have cash available if you’re willing to pay for it. Therefore, futures prices always reflect current prices, not the expectation of the volatility in future prices. All you’re looking at, essentially, is the yield curve, not the oil market volatility.

    I’ve seen people make this mistake before about futures markets, thinking that they divine what the market thinks the future price of a commodity is, and it has become a bit of a pet peeve of mine. The assumption is dead wrong. The futures market is essentially a very leveraged way to buy a synthetic representing the commodity itself as it is priced right now given certain interest rate expectations. If you want to see what the price of insurance on the commodity is, you look at the options.

    If you really want to look at the market’s expectation on volatility, look at the options market, not the futures market. That tells you what people believe the volatility is going to be — you can get the implied volatility by reversing the Black-Scholes model, backing out the volatility given the price and interest rate instead of getting the price given volatility and the interest rate.

  • Perry E. Metzger

    I realized some readers might not understand what I just said, so I’ll expand slightly.

    Imagine a futures contract for widgets. Say the impossible happened and that although people were willing to pay $10 for delivery next week they would pay $100 for delivery in one year. A smart trader could trivially buy widgets today at $10 and sell contracts at $100, locking in a future $90 profit. (Exercise for the reader: figure out what you would do if widgets for delivery next month were $100 but in a year were $10).

    Thus, arbitrage will make the prices of all futures contracts come in to line with the current spot price.

    This isn’t quite correct, of course. We ignored two important factors. First, if you’re buying something today to fulfill a contract in two years, you need to consider how much the capital committed over that period will cost you — whether it is in lost interest at the risk free rate of return or in interest payments, it is pretty much the same thing. Second, you will need to store the widgets over the period, and storage costs also need to be accounted for. Those factors alter your expected profit.

    Also, interest rates are not the same for six months or three years, and neither are storage costs — interest rates especially follow a curve known as the “yield curve” in bond market parlance.

    Therefore, the future prices will form a characteristic curve that reflects the current price of the commodity with adjustments for the current yield curve in the bond market and the cost of storage and similar incidentals in time. (This curve can be quite interesting — in some markets, such as the gold market, the curve effectively allows you to read out the rate of interest on gold loans even though there isn’t a modern market for loans of gold per se.)

    As I said in my last posting, I have a big pet peeve about people looking at futures prices and announcing the market things the price of X will be stable for years. This sort of claim happens all the time, and it is always made in apparent ignorance of the fact that futures markets don’t work that way and their prices don’t tell you that sort of thing.

    As I also said, the options markets are quite different, and if you believe the Black-Scholes model, you can even read out the “implied volatility” of the market based on the options prices. They do have one flaw, though, for the prognosticator, which is that you sometimes can’t find freely traded options on the thing you want to assess the volatility of (and even if there are options, often the ones for very far off dates aren’t traded because they would be too outrageously priced even at reasonable volatility levels…)

  • ChrisA

    Thanks Perry, I had been making the mistake you warn against, and your explanation showed me why I was wrong. Putting your words another way, it makes sense that the most important indicator of future prices is todays price, since if sellers thought they would get higher prices in the future they would wait for them.

    Other indicators that we are not near peak oil is the share price of existing big oil companies, if the market thought we were near peak oil their share p/e ratios would be dot com like, they are actually at a discount to the market p/e ratio.

    The peak oil debate is really just Malthusiasm dressed up. The proponents say because I can’t prove that humanity will find oil beyond today’s known reserves, then we must assume that they will not be found. But history’s guide is always that the malthusians are wrong. Even the very mature science and engineering of oil extraction is no-where near perfection, and we have barely started on improving alternative energy technologies.

  • Perry E. Metzger

    I’m not sure that I would be so quick to dismiss “peak oil” per se. We passed peak horseshoe production long ago, didn’t we? Products have life cycles, and eventually we may no longer be producing much oil. Indeed, just considering fossil oil, we might very well hit peak production even if we would “want” to keep extracting. You cannot continuously remove a finite amount from a fixed finite pool for an infinite period.
    It is pretty clear that, at some point, oil will cease to be quite so important from the point of view of economics, or if it continues to be important (perhaps because of the needs of the petrochemicals industry) fossil oil will run out (or become too scarce for feasible extraction) and be replaced with synthetically produced oil if there is really such a pressing need for alkanes for whatever reason.

    That does not mean, however, that any crisis is looming. I would say one should carefully distinguish the doctrine that oil in the ground will eventually become scarce if we keep using it (probably obvious) from the notion that we would face some sort of societal problem as a result of this.

    Quite clearly no such crisis would appear — there is plenty of usable power around, and as the market starts to send price signals to producers and consumers, they will quite naturally move to the use of other power sources. The high prices in the markets for fossil fuels are already driving vast investment in substitutes, including (IMHO) the most obvious ones, nuclear and solar.

    The real issue is not “peak oil” — it is the misunderstanding that assumes that the world is static and that markets will simply “drive into a brick wall” unless caring central planners carefully steer them away. However, markets do, in fact, have a way of planning without planners — price signals — and price signals will be just fine in the oil markets if we don’t interfere with them.

  • http://profile.typekey.com/halfinney/ Hal Finney

    Perry – I’ve seen that argument and it does make sense. Particularly for a storable commodity like oil, futures prices cannot get too far from today’s spot prices.

    At the same time, it is undeniable that you can buy and sell oil today for future delivery at the contract price. If a trader thought that oil in 2007 was going to be $80 rather than $65, he would buy contracts for 2007 delivery of oil at $65 with the expectation of selling them for a profit at $80 next year. If this $80 price were the market consensus, everyone would buy up those contracts until they reach $80. Therefore futures prices cannot be different from the market expectation of future prices.

    The way I resolve this paradoxical condition (that futures prices must track today’s spot prices, while they must match the market expectation of future prices) is to say that today’s spot prices must reflect the market’s expectation of future prices. If the market really thought oil would be $80 next year, it could not be selling for $63 today. It would be selling for closer to $80 today on the spot market, because sellers would be holding it off the market and waiting for the expected high prices a year from now.

    Therefore my understanding is that both views are correct, but most importantly for the purposes of gathering information about what will happen, futures prices do reflect the market consensus on future prices.

  • Perry E. Metzger

    Hal, you began this by posting the price curve and asserting that it tells us something about market expectations about peak oil. That is, unfortunately, not correct.

    Certainly today’s spot price tells you a bit about what the market thinks the future holds, but it tells you less than you would like. In particular, near term supply and demand of necessity is the major influence on the spot price, not long term supply and demand. People have strong time preferences.

    The fact that the market has the word “future” in the name makes people think it is about future prices, but the futures market is not about future prices, it is about locking in spot prices for future delivery while putting down only part of the full price, and there is no information to be derived about volatility at all from looking at the whole curve you presented. The future’s market is a just way for people to buy things for future delivery without having to put down the entire cost in advance. It is not a prediction market.

    Your original post shows the whole curve, of course, as though the curve were meaningful for figuring out expectations about prices, and the curve is not meaningful for that. Option prices are meaningful for this, not futures prices.

    If you want proof of this, look at as many futures contracts as you can, and explain to me why it is that they all reflect the current interest rate yield curve. Surely if there was some element of prediction involved they would not all have the same curve. Surely every single commodity can’t have the same expectations about price volatility. At the same time, however, option price implied volatilities have wide variation.

    In the spirit of “overcoming bias” and such, I would suggest re-do the exercise in a way that is actually meaningful.

    I’ll repeat, there is a market that tells you about expected future prices, and it is the options market. Options prices reflect market expectations about price volatility and can be used to assess said expectations. I would strongly suggest that what you do is repeat this exercise, but post instead the prices and implied volatilities for options on oil futures with a variety of future expiration dates. That information will be much more interesting from the point of view of answering “what do the markets think will happen to oil prices.” The needed information is of course freely available so it should be easy to get your hands on it.

  • http://profile.typekey.com/halfinney/ Hal Finney

    Perry, here are a couple of other futures strip charts I just made. All reflect Friday’s closing prices. First gold:

    http://alumnus.caltech.edu/~hal/gc20061222.png

    As far as I can see it’s a perfectly straight line. It looks nothing like the oil chart.

    Next, natural gas:

    http://alumnus.caltech.edu/~hal/ng20061222.png

    This very clearly shows the market expectations of future prices. It is up and down like a roller coaster. The market expects gas to be expensive in November 2010 and cheap in May 2011. This is of course a reasonable expectation, but the point is that the price chart does in fact reflect the market expectation of future prices. If it were just today’s price plus some interest rate premium, it wouldn’t go up and down seasonally like that.

    One thing I’ve never seen explained is why the oil curve has the shape it does: up and then down. Has it always had that shape? Or is it a recent development? Gold looks completely different: it’s a straight and steady upward slope. NG does have a similar shape to oil if we just look at the trend and ignore the seasonal fluctuations, up and then down.

    Does this mean that the market actually expects energy to get more expensive for a year or two, and then start getting cheaper (perhaps as high prices motivate new investments in production, which take a few years to come on-line)? Or is there some other explanation? And why do other commodities look so different?

  • Errol Alvey

    I would like to give ChrisA some food for thought, regarding his/her assertion that “there is plenty of useable power around”.

    There really is nothing else in the energy world quite so versatile, portable, energy dense, and inexpensive as petroleum. Petroleum is what economists refer to as a “subsidy”, that is, roughly something for nothing. When a petroleum deposit is first tapped, it is almost always naturally pressurized and it isn’t even necessary to pump it out. Once a well was drilled in West Texas, all that was necessary to get cheap, concentrated solar energy was to open the wellhead choke! This cheap power was then used to build the very complex and expensive oil production/refining/distribution system we now benefit from. Petroleum provided the financing and concentrated energy to construct its own infrastructure.

    Chris mentioned nuclear and solar as substitutes, so I will discuss each in turn. Nuclear power requires very substantial fossil fuel inputs. The concrete for the massive containment building CAN NOT be made with nuclear heat or electricity. The mining of uranium CAN NOT be performed with nuclear heat and only with great difficulty with nuclear-derived electricity. Uranium is also a finite resource; the price signal has been disconnected for the past 10 years or so by the sale of U235 and Pu from decomissioned Soviet weapons. So here we have a complex, capital intensive system that does not provide the energy for its own construction. No subsidy here!

    Solar power (this includes wind power, which is atmospheric convection) is very diffuse, intermittent, and unevenly distributed. Diffuse means that solar will be a very poor subsitute for activities requireing energy density (think cement production) or portability (think transport). Intermittent requires little comment – think cloudy days and days when the “wind” is actually just a breeze. Unevenly distributed is also pretty obvious – Gemany and Great Brittain are notoriously cloudy places. Often times the best places to collect solar energy are far removed from the places where the power consumers are. All of this adds up to an expensive, intermittent system that will not provide the energy for its own construction. No “something for nothing” here, either.

    Finally, please understand that the “price signal” for petrolueum has been intentionally obscured by central planners in OPEC. The Kingdom of Saudi Arabia in particular has been very shrewd in keeping the price of petroleum in a range that maximizes their return on investment while discouraging the development of alternatives. When KSA starts their inevitable decline (while China and India continue to ramp up consumption), we will find ourselves trying to build alternative energy infrastructure under horrific time pressure and using much more expensive fuel.

    In summary, alternatives are poor substitues for petroleum (or they would be in wide use already), and price signals have been distorted. The time to make massive investments in conservation and alternatives is now.

  • Nick Ramm

    Hal –

    Perry is right that futures prices do not reflect futures expectations. The reason for the funny looking futures strips is because of “convenience yield.” This measures how favorable it is to hold an asset physically rather then hold a futures contract on the same asset. A rise in the convenience yield therefore lowers the futures price. Since when inventories are high the risk of shortage, and hence advantage to holding the commodity over the derivative are low, the convenience yield tends to be low when inventories are high. The “convenience” of holding physical natural gas as opposed to a futures contract is lower during the coldest months since that is also when inventories are at their highest. Likewise, both oil and natural gas show a downward tendency in their futures prices over time as the risk of a shortage that far out in the future is greater. Most gold is not consumed, but rather held as an investment asset. Therefore the upward sloping futures price reflects foregone interest made by holding gold rather then selling immediately along with storage costs.

  • Perry E. Metzger

    Hal;

    Congratulations. You picked two commodities for which the futures prices will always look odd. I mentioned one of them explicitly already, but you didn’t pick up on that I see.

    First, gold. If you re-read my earlier messages, I specifically said that the gold chart will not look like other ones. You will note that I already mentioned the reason why gold has that line. It is telling you the implied interest rate on gold loans. Exercise for the reader: what do I mean by “the implied interest rate on gold loans”, and why should gold be different this way?

    As for natural gas, it has an interesting feature to it, which is that it cannot be stored easily, unlike oil or iron. See if you can figure out the pattern in the natural gas prices. There is one. (Hint: Notice that the curve is self similar if you translate it by one year, with an odd inflection point in September. What happens to temperatures in September? Is this roller coaster at all “random” as that term is usually understood?)

    As for the reason for the curve on oil prices, it is because of the yield curve on interest rates (and also the cost of storage). Have a look at the bond yield curve, and you will understand most of the oil curve.

    I would suggest looking at a more normal commodity if you want to see a curve exactly like the one for oil. Try aluminum.

  • Perry E. Metzger

    Errol;

    I think your view of both nuclear and solar is very simplistic.

    There is enormous net energy out from nuclear power plants, and ignoring any sort of subsidies etc. the price of nuclear generated electricity is lower than that even for coal. If we had nothing but nuclear plants, we would find ways to get concentrated fuels for purposes like mining and powering aircraft by synthetic means. The amount of natural uranium we have around, properly husbanded, is enough to power our civilization even given reasonable rates of growth for longer than the planet will survive. See John McCarthy’s “Sustainability Page” on the web if you want to see the calculations to prove that.

    As for solar, it is not “diffuse”. It is perfectly reliable, too — the solar constant in earth orbit is a nice 1400W/m^2 (at the surface it is more like 1000W/m^2 and less reliable) and with easily seen technologies we should be able to get upwards of 60% of that. The answer to the fact that the sun is not always shining at the surface of the planet is not to build your system so that it depends on the sun always shining, which is easily done with both orbital power arrays and various forms of mechanical and chemical energy storage. It is true that the efficiency of storage is not always optimal, but that is easily made up for in how much solar power is available. Right now the cost is prohibitive, but that is largely a manufacturing issue. The current reason photovoltaics are expensive is the high current cost of producing monocrystaline Si — this will not be a barrier over the long run. Even without any fundamental changes in the techniques used to produce monocrystaline Si the quantities being made are currently orders of magnitude below the production of, say aluminum, and economies of scale with nothing but current production techniques would be sufficient to make solar economical. When molecular manufacturing becomes practical the price would fall through the floor, even if we just used Si, as Si is one of the most abundant elements on earth.

    The only real reason to be interested in petroleum long term is that alkanes such as octane are a very high density means of energy storage. However, as I noted, it should be feasible to synthesize them for applications such as air transport where storage density is crucial.

  • michael vassar

    Perry, do you know the approximate efficiency and capital requirements of simple methods for synthesizing hydrocarbons, such as electrolysis followed by reduction of CO2 with Hydrogen?

  • Perry E. Metzger

    Right now, all methods of synthesizing long chain alkanes are uneconomical. However, that is unlikely to remain the case permanently. Biological systems efficiently synthesize long chain substituted alkanes, to whit, fatty acids, so in principle we should be able to ultimately harness those same synthetic pathways or find even better ones.

  • michael vassar

    Of course they are uneconomical now. We still burn hydrocarbons for energy so obviously it’s not economical to convert electricity to hydrocarbons. However, many sources of energy in the future may be better for making electricity than for making hydrocarbons. If electricity was X times cheaper than oil, on a joules per dollar basis, we could economically convert it to oil. I want to know X. A good starting point is to ask the energetic efficiencies of electrolysis and of reduction of CO2.

  • http://profile.typekey.com/robinhanson/ Robin Hanson

    This forum is about Overcoming Bias (TM). Since the topic of peak oil is controversial, it is plausible that it might exemplify important biases or ways to correct bias. But I haven’t seen that very clearly in this post or these many comments. It is good that we just don’t talk abstractly about bias, but grapple with concrete examples. But let’s try connect those concrete examples back to the subject of bias; I don’t want this to just be a general forum for discussing controversial issues.

  • ChrisA

    Hopefully this is relevent to the subject of bias – what interests me about the peak oil debate is how often the debate on future shortages of various commodities endlessly re-occurs. This started with Maltheus, but there are lots of other examples since (most recently the Limits to Growth report, Paul Erlich and the peak oil protagonists). The case is always presented very strongly; with little doubt as to the imminent terrible outcome should the chosen policy prescriptions not be followed.

    Of course, so far, none of these terrible outcomes has happened, even with the proposed prescriptions not being followed. This is probably a good thing as the prescriptions generally were very expensive or required significant sacrifices. But the failure of these previous predictors does not seem to deter the current lot, even though their arguments seem to me to basically be the same.

    Perhaps the attractiveness of the scenario of imminent commodity shortages is somehow related to a folk or even genetic memory of the days when life was largely agricultural and this tendency was a useful reminder to prepare for lean days even in a time of plenty.

  • Perry E. Metzger

    I think it is important not to bias oneself against “peak oil” by assuming that such a prediction implies a Malthusian or Erlichian outcome.

    Many minerals have, in fact, been exhausted over the millennia, but mankind has survived anyway, because we can come up with perfectly reasonable substitutes. For example, natural cryolite deposits more or less ran out decades ago, but we now use synthetic cryolite in aluminum smelting so the world hasn’t really noticed.

    Now, the “bias against Erlich” would hold that it is impossible that natural cryolite could have run out, but indeed it did. It just wasn’t a catastrophe for the industry. The Erlichian view would have predicted catastrophe, but one must not “overcompensate” by assuming that resources cannot run out.

    Indeed, I will point out that groups of humans have gone insane at times and destroyed the basis for their own survival — see Easter Island, where the obsession with erecting monoliths lead to the elimination of all trees on the island.

    The “peak oil” meme should be analyzed in two distinct parts. The first is, “is it credible that oil would eventually run out if extraction continues at current rates”, and the answer is an obvious yes. When it would run out, I can’t say, but we’re clearly removing it faster than natural processes re-create it, and mathematics tells you what the ultimate destination of those curves is — at some point, extraction of fossil oil will of necessity decline. The second part is “should we, on a policy level, care?”, and there I think the answer is “no”, because there are adequate substitutes, and the market can deal with finding them.

    A possible third question is “how soon will we hit the peak on the production curve, because I want to know how to invest”, and there I can’t answer. Every entrepreneur must figure that out on their own.

  • John D

    Consistent with Hal Finney’s treatment in this post …

    Here’s how Alan Greenspan looked at crude oil futures at the beginning of an April 2004 speech: “The dramatic rise in six-year forward futures prices for crude oil and natural gas over the past few years has received relatively little attention for an economic event that can significantly affect the long-term path of the U.S. economy. Six years is a period long enough to seek, discover, drill, and lift oil and gas, and hence futures prices at that horizon can be viewed as effective long-term supply prices.” (http://www.federalreserve.gov/boarddocs/speeches/2004/20040427/default.htm )

    In October later that year Greenspan noted the earlier behavior of futures: “Between 1990 and 2000, although spot crude oil prices ranged between $11 and $40 per barrel for WTI crude, distant futures exhibited little variation around $20 per barrel. The presumption was that temporary increases in demand or shortfalls of supply would lead producers, with sufficient time to seek, discover, drill, and lift oil, or expand reservoir recovery from existing fields, to raise output by enough to eventually cause prices to fall back to the presumed long-term marginal cost of extracting oil. Even an increasingly inhospitable and costly exploratory environment–an environment that reflects more than a century of draining the more immediately accessible sources of crude oil–did not seem to weigh significantly on distant price prospects.” (http://www.federalreserve.gov/boarddocs/speeches/2004/200410152 )

    To quickly relate this discussion at least tenuously to the topic of bias (though bias here could be a risk premium consciously embedded into the market price)…

    Futures as a biased predictor of spot in crude oil:

    “This paper presents some empirical evidence on market efficiency and unbiasedness in the crude oil futures market and some related issues … The evidence suggests that futures prices are neither unbiased nor efficient forecasters of spot prices” (http://www.sciencedirect.com/science/article/B6V7G-458XJS2-69/2/9696bd7247997cb4fef83a6ece71a9c1 )

  • http://profile.typekey.com/halfinney/ Hal Finney

    Perry, here is a curve I made today for Aluminum based on prices just before closing:

    http://alumnus.caltech.edu/~hal/al_20061226.png

    It looks pretty different from oil to me. First, oil (after the first month) rises then falls, while aluminum (after the first month) only falls. Second, if we restrict attention to the time frame up to the end of 2008, which is as far as the Aluminum futures go, oil rises to a peak during that time frame, while Aluminum falls. So during this time period the two commodities are going in opposite directions.

    It is important to understand the degree to which futures prices reflect market anticipation of future spot prices. Certainly futures markets are more socially useful if they can be considered to apply this function, and in that case they offer great value to those of us seeking to Overcome Bias.

    I have yet to see an explanation of how a futures contract price can substantially diverge from the market consensus future spot price. Clearly in that circumstance, the average trader can take a position with an expected profit, which should not be possible in an efficient market. Now there are costs to such a position: transaction costs, but they are not too large; foregone interest, but commodity contracts require only a small margin payment so this is small compared to other investments; and risk, but this applies equally to long and short speculators, and most markets have few hedgers, so it is not a major factor.

  • Perry E. Metzger

    Hal;

    As I explained earlier, because you can store a commodity, the price for the futures contract tracks the current spot price modulo the interest rate and storage costs, not the future price. You can learn about this in most books on futures contracts. To someone with an economics background, this sort of argument should be straightforward. Naturally, the current best guess about the future price does alter the current price, but it tells you nothing about the beliefs about expected range.

    Were my statement untrue, then there would be an amazing opportunity for arbitrageurs — they could simply buy vast lots of commodities at spot, and lock in profits in the futures market with virtually no risk. You can choose not to believe me if you like, but then I’m afraid you would have to explain why it is that people could not arbitrage the price on a contract closing in two years against the current price, which is essentially what you would have to claim in order to claim that the price on the two year contract does not reflect the spot price. Indeed, were my statement untrue, why aren’t you yourself out there making a fortune arbing the spot market against the futures?

    As for the aluminum curve you posted, it looks rather strange to me — the two completely linear domains are very odd, as is the denomination on the vertical scale.

    Anyway, I’ve said as much as I think is productive on this topic.

  • http://profile.typekey.com/halfinney/ Hal Finney

    Perry, what you say makes sense and I agree with it. However at the same time I think the point I made is valid as well, that there is an arbitrage opportunity if futures market prices are out of line with market expectations of future spot prices. I believe these two points of view are compatible, as I argued above. And as the Greenspan quotes that John D found illustrate, the shape of the futures strip does contain information beyond what is found in current spot prices. Back in the 90s it apparently had a different shape from today, and the question is what this tell us about the future.

    I found an article that discusses these points in some detail, and includes a graph comparing the shape of the curves for different commodities:

    http://www.bankofengland.co.uk/publications/quarterlybulletin/qb060105.pdf

    See Chart C in particular which illustrates an oil price curve similar to the one I showed above, along with gold and silver, and non-monetary metals aluminum and copper. The gold and silver go up in a straight line for the reasons you have explained, but aluminum and copper both fall, again similarly to the aluminum curve I made today.

    They do make an interesting point for why oil futures prices in particular might overestimate spot prices. The idea is that oil speculators with long positions are getting insurance, since if oil prices rise it will hurt the economy and could reduce speculators’ income. By taking a long position they offset that risk, while short speculators take on more risk. This puts upward pressure on the futures market price and might cause it to be higher than expected spot prices.

    The specific example they give shows a Reuters survey of economists predicting under $45 in 2010, compared to the oil futures price of over $60. Maybe I spend too much time with Peak Oilers but that $45 prediction sure looks low to me, and it’s hard to credit it as superior to the market price of $60+.

  • http://www.odograph.com odograph

    I was intrigued by Peak Oil as an idea 18 months to 2 years ago. It seemed such a strange return to the fuel crises of my youth. I started following the energy news (and the more reasonable of the blogs) more closely, read a few (of the more reasonable) books, and came to broadly accept something like Peak Oil. I could say I fuzzily expect a peak, or something fuzzily like a peak.

    What I think sources like Overcoming Bias teach me though is that I can’t really say much about what that peak will look like. I certainly can’t saw what the world after the peak will look like. There are way too many questions between here and there. We don’t know which technologies can be improved, and we don’t know what social trends will predominate.

    Anyone who describes a post-peak world is layering a considerable set of assumptions in order to get there. As an example, how many optimists assume that energy will be cheap and plentiful after peak oil? How did they get there? Is there a rational path to that (or any other) outcome? Or is it an assumption that the universe owes us cheap energy?

    This is an interesting story, and the continuing human reaction as it plays out is not a small part of the drama.

  • Jon

    Hal – re your Dec 24 post

    “One thing I’ve never seen explained is why the oil curve has the shape it does: up and then down. Has it always had that shape? ”

    No. It hasn’t. A great deal of what you’re seeing is a result of spreads calling the commodity to the marketplace – or paying the producer to keep it off the market. (Yes, there are many, many other factors – the answer below is simplified)

    I’m a grain hedger, so I’ll explain with grains. The spread between Jan beans and March beans closed at $.15 yesterday. This means, to me in my position, that the market will pay me $.075 per month to hold my grain from the cash market. Essentially, the market is paying me to keep grain in storage – the deferred price is higher and the strip slopes up.

    However, when the commodity is in demand in the market place, an inverse occurs – the front price is ahead of the deferred price and it costs the producer/hedger to hold onto the commodity. If Jan beans were $.15 more expensive than March bean, then it’d cost me $.075 per month to hold the beans in my bins. Now – there may be reasons for me to do so, however, the market is clearly telling me to bring the bean to the cash pipeline now.

    Back to oil. The strip tells us that there is more supply than demand – the market effectively pays the producers to not bring it to market by the amount of the spread differential. However, why does it then do inverted from the contango? The market “thinks” that there will be a time when demand will outstrip supply and thus, it will pay the producers to bring the oil to the market rather than leave in storage.

    One additional factor – the effect of the “long only” commodity funds. Big investment pools that buy commodities as an “investment”. I know they’re always long, the locals know they’re long and we take advantage of their rolling by front running the spread driving it out before they do. This adds an additional contango factor and whole additional level of complexity – and interest.

  • Mr. InternationalParliamentNow

    Look at the forest not the trees:

    1.everyone is having to go much deeper for oil.

    2.The gulf state nations may really have reasons for wanting nuclear power other then dropping the bomb on Israel.

    3. Demand for oil will increase with the further industrialization of china/india etc.

    4. Why go after the Anwar to drill for oil if we don’t have the need. Please don’t talk about our dependence on foreign oil as the reason as what would a 10pct increase home grown oil do to our purchasing from the Saudi’s? (Think about it)

    5. Within how many years will it be before there are mass shifts in population due to the effect of rising seas. (Don’t believe me:look at the ice caps and glaciers) Oh but that’s not Global warming? Look at the ice caps and glaciers (Who cares what caused it: Look at the ice caps and glaciers!!! This will create a further capital drain, perhaps mass projects to save florida, the netherlands, california coastline, etc.
    (Don’t believe a word I say – just keep watching.) Seeing is believing.

    6. WAR: As resources become scarce as we begin the shift away from Fossil fuels man, we are already gearing up to make sure we own all the oil that’s left.

    7. We are building a wall around our Country to keep out the riff raff. If it gets bad here imagine how bad it will get elsewhere.

    8. Our national guard and (unbelievable) state guard are being called up to their third extended tours in Iraq. That is the National guard – remember – they used to be weekend warriors. Go to Iran? Without a draft?

    9. China is training pidgeons to fly under electronic direction. How bout that bird flue? China has shot a sattallite out of the sky. China is in the midst of a military buildup that we don’t talk about. Russia, is not so benign anymore and is starting to grimice at its neighbours.

    10. And drum roll please – The Vice president and President – I am sure – outed an undercover CIA agent who’s husband was blowing the whistle regarding the real intelligence on the war in Iraq.

    Well that about wraps it up – is anyone awake out there?

    Oil at $100 plus- coming soon

  • http://profile.typekey.com/halfinney/ Hal Finney

    No offense, but that looks like a lot of “trees” to me. That is, you are analyzing the situation by studying detailed facts. My goal is to step back and look at the big picture without getting distracted by the facts(!) – to, as you say, see the “forest”.

  • MR.InternationalParliamentNow

    you are lost in a world of your own analysis:

    “Back to oil. The strip tells us that there is more supply than demand – the market effectively pays the producers to not bring it to market by the amount of the spread differential. However, why does it then do inverted from the contango? The market “thinks” that there will be a time when demand will outstrip supply and thus, it will pay the producers to bring the oil to the market rather than leave in storage.”

    All meaningless. While you are fiddling Rome is burning. Bigger things are happening. Don’t look at the facts . Everything will be okay. 100 dollars a barrel and above. Want me to predict a NLT date?

    I don’t need a big book of formulas to prove my hypothesis.

    But it’s not the end just a major transition in the evolution of civilization

  • Taemojitsu

    A very confusing conversation. Giving Perry E. Metzger the benefit of doubt perhaps he and Hal Finney were talking about two separate things. “Volatility” to the average uninformed person like me could either mean the change in the price of a commodity from now until the settlement time, or it could mean uncertainty in the future price. Without having read the followup entry, I think there are several conclusions one could have made from this discussion:

    1) The original post unconsciously assumed that oil spot prices will continuously increase, at whatever rate, once oil production peaks, in other words that other factors would not affect that upward trend
    2) The expected spot price of a commodity in the future is affected by exactly the same factors as the price of a futures contract for that commodity, so they are the same thing
    3) Various complicated things affect the prices of relatively low-priced metals that are not meant to be stored in bulk, such as aluminium, which may cause downward trends contrary to interest rates/inflation
    4) Highly valued commodities like gold will have upward-trending futures as they have relatively low storage cost so are not sensitive to transient supply or demand issues around the world
    5) the downward bend in oil was due either to an initial partial payment by the buyer (and thus loss of interest), or an expected decrease in demand due to alternative sources of energy, or (most likely) to a reduction of supply issues that led to the predicted spike; but was most definitely not due to any “long-term storage costs” which only affect the linearity of a futures market with respect to its current price. For a commodity with low storage costs future demand will cause a large shift in current price which reduces the future spikyness, while for high storage prices it is smoothed mostly by changes in demand or supply which leads to more spikyness such as in natural gas.