May 21, 2006
Natural gas and crude oil prices
Can natural gas fall this far without oil prices moving more?
A barrel of oil has about six times the energy content of a thousand cubic feet of natural gas. Although they're far from perfect substitutes, some firms can switch between oil and natural gas depending on price and availability. Longer term, consumers or firms making a fixed commitment to oil or natural gas will of course look at the relative prices. Thus one's first guess might be that, over the long run, a barrel of oil should sell for something around six times the price of a thousand cubic feet of natural gas.
The graph below shows that this has been a pretty fair description over the past six years. Up until the last few months, that is, which have left natural gas selling for about half its value from December, while oil is little changed.
If the historical correlation holds up, something's got to give here. Strikes me as another reason to wonder about oil above $70 a barrel.
Posted by James Hamilton at May 21, 2006 07:45 PMdigg this | reddit
For what it's worth, these guys--Raymond James and associates--argue that the ratio should be 3-to-1 instead of their earlier report of 5-to-1 because of efficiencies that exist in the use of natural gas.
I don't have the volumes that can be switched from one source to another. But the 3-to-1 ration argues that (a) there is greater incentive to use gas but (b) there has almost always been an incentive to use gas and therefore it seems that everyone would be using gas instead of oil anyway. Even the high natural gas prices of late 2005 would have been a cheaper source of energy than oil.
If so, then the downward pressure on oil prices created by natural gas prices (and fuel switching) is true in theory but in practice may not amount to much.
Irrespective of that, I don't see much upward pressure on oil prices right now other than unforeseen calamity.
Posted by: T.R. Elliott at May 21, 2006 10:37 PM
Just a note from someone in the power biz... Nat gas and transportation fuels may have parted ways for a reason.
There was a glut of natural gas-fired power plant construction in the US, starting in 1998 or so, and ending with the bankruptcy of Calpine in 2005. This construction boom drove NG demand (and prices) through the roof. In fact, in 2002 for the first time ever, daily NG usage peaked in the summer months (for power generation) rather than during a cold snap (for home heating).
Since that time, many of these facilities became uneconomical to run, due to exceedingly high nat gas prices and low energy prices. These factors were obviously not foreseen when the facilities were financed and built.
These economics are most clearly observed in areas with more diverse energy supplies (hydro, coal, nuclear) and that is where you see idling of the new gas-fired plants.
I simply do not see the same curtailment of demand for crude oil, nor the diversity of replacement fuels for transportation.
On the other hand, there is a relationship between the two trends that cannot be denied. Here is the question for you: In light of the consistent upward trend of *both*, why do you believe that the *downward* trend of NG is giving you the correct signal, rather than crude oil?
Posted by: Idaho_Spud at May 21, 2006 10:48 PM
Excellent graph and tantalizing question about $70 oil. Now, if those 2 spikes in ngas weren't there, it would even be more clear cut that something/body is misbehaving in the last 6 months.
Does looking at a wider time frame confirm this view?
Posted by: calmo at May 21, 2006 10:55 PM
(remove at to reply by email)
Calmo and TRE: excellent points.
I think inherently natural gas is a much more seasonal fuel than oil. Because of its role in heating (winter) and now power generation (summer).
On the latter, gas powered plants tend to be used at 'mid merit' and 'peak load' not baseload, ie they are only run to supply the last 30% of so of power demanded, so the demand is spikey. Many commercial users of gas can also schedule production when prices are too high. by contrast, there is almost no demand for oil for power generation (in the US).
Therefore gas prices tend to peak just before winter starts, and may also now show a summer peak (before the peak power demand).
Gas is also inherently far less flexible. You cannot change the proportions of eg jet fuel v. gasoline coming out of the cracker-- when you produce gas you produce one commodity, that you can only deliver to consumers where there is a pipe (LNG isn't big enough to affect the US market price).
My gut is the right comparison is with gasoline (also a summer peak fuel) rather than crude oil per se.
So the gas price signal may not be a useful signal for anticipating the crude oil price.
Posted by: Valuethinker at May 22, 2006 06:02 AM
Does natural gas use just have higher price elasticity than liquid fuels?
Posted by: odograph at May 22, 2006 06:18 AM
These claims can be tested by observing whether r has changed over time.
Posted by: Robert Schwartz at May 22, 2006 07:25 AM
Is storability an issue? As an ignorant layman, I have the vague impression that petroleum would be easier to store than gas. If so, the arbitrage might be stronger in one direction than in the other. When gas is underpriced for the long term, you would not get the same sort of price correction that you might get when oil is underpriced.
Posted by: knzn at May 22, 2006 07:59 AM
I actually think (don't know) it is the other way around. Gas is easier to store. You either pump it out of the ground (ie open the valve and let it flow) or hold it there, or in storage tanks.
There is little or no processing that it undergoes before it goes to its end customer, and you can regulate production exactly to meet demand.
The problem comes, as there was last winter, when there is not enough gas in storage or production to meet threatened demand, and so the price spikes. This is a particular problem if a lot of gas fired electric generation capacity is also running (winter heating peak can coincide with winter lighting peak).
Posted by: Valuethinker at May 22, 2006 08:46 AM
As Valuethinker noted, there is considerable seasonality for gas. Current prices are low for two reasons: 1) little weather-related demand with no heating and no air conditioning at this time and 2) an overhang of natural gas storage due to mild winter weather that consumed less stored gas than usual.
An important characteristic of natural gas storage is that it is rate-limited as is its transportation. One can only inject or withdraw gas from storage as certain flowrates. Oil can move much more quickly in and out of storage tanks on a BTU basis plus liquid oil storage is cheaper and more flexible. Gas storage is usually located to serve the end-users.
In the energy industry, gas and oil have had a declining correlation over the years due to increased pollution control regimes. Oil is much more focused on transportation these days. Switchability is more and more a thing of the past. That trend could turn around due to future price disparaties but it would be a slow and laborious process due in part to bureaucratic resistance.
I'll bet that natural gas will climb against oil this summer due to electric production demand, especially if it gets hot.
Posted by: Joseph Somsel at May 22, 2006 09:26 AM
I, too, am uncertain that crude and gas necessarily should move in lockstep. Vanishingly little crude demand is liable to fuel-switching, so I simply don't expect much relief from low NG prices.
That said, I hope Mr. Hamilton put on this trade (short crude/long gas futures) at the open today. He would appear to be up over 4% in less than a day's trading. Nice work!
Posted by: wcw at May 22, 2006 09:58 AM
Excellent post & discussion. Thanks.
Posted by: algernon at May 22, 2006 12:35 PM
I think gas futures rose today on the release of the Hurricane forecast. No surprise to anyone monitoring water temperatures in the North Atlantic and the Gulf, which are already substantially above normal.
As pointed out above, gas demand is at its low in the shoulder seasons, and will rise as air conditioners come on full time in the next few weeks. There is a substantial overhang in storage, however, and it will take a hot summer to eat this up. This is what is being forecast.
Long term fundamentals for gas are strong as reserves continue to fall, and exploration companies are drilling a record number of wells to increase production. Unless and until LNG becomes a factor in N. America (2009 or later)demand will continue to rise and supply is constrained.
Posted by: David Baskin at May 22, 2006 01:46 PM
Oil is pretty much used exclusively for transportation fuel. 75% of the barrel goes to gasoline, jet, and diesel and most of the rest would if it could.
There may be some minimal switching users left, but not many. I suspect the one's that are left probably run gas most of the time and run heating oil in the gas price spikes. They sure aren't going to be shedding oil demand now. And they're small ebough now (a lot has happened since 1994!) that their demand is probably noise in the system. No doubt this spread will drive more ethane capture/less naphtha cracking to the extent petchem plants have this flexibility as well. This may be the bigger swing.
If you really look at it, the price setting mechanism for oil has changed in recent years. During most of the period up until recently, oil prices have been largley a product of OPEC's ability to manage its excess capacity.
Now, it appears more like oil prices are rising to the point where enough worldwide demand is destroyed to balance with available supply. This requires making drivers stay at home.
Nat Gas, on the other hand, is largley a domestic/canadian product. High prices have resulted in a drilling boom. Supply has out-grown demand.
Posted by: dis at May 22, 2006 02:06 PM
I am unable to detect from the price graph spanning the last 7 yrs, the seasonality of ngas. Unlike Value who spots it right away:
Therefore gas prices tend to peak just before winter starts, and may also now show a summer peak (before the peak power demand).
If it were true that winter use or summer use was disproportionately seasonal, wouldn't it have shown up in higher prices during those periods? (Is this a technical/accounting question about how/when prices are registered?)
If it is true that ngas overcapacity is to blame (as per persuasive Idaho) I am mystified that current gas prices have not had any downward impact on oil prices.
Also not seen (by me) on the graph is Idaho's:
Since that time, many of these facilities became uneconomical to run, due to exceedingly high nat gas prices and low energy prices.
I am unable to disentangle this (relatively) high nat gas prices from the low energy prices --esp in the last 6 months.
How much does it cost to convert your old truck to LNG? Something looks fishy, no?
Posted by: calmo at May 22, 2006 02:31 PM
On Idaho's point he means 'electric energy prices' which are not on the graph. The 'spark spread' between gas prices and electricity prices has been too small for most gas fired gencos, the last few years.
The seasonality is relative to baseline prices that year. I think if you look back, you will
see distinct spikes in gas prices just before the heating season starts.
From what I understand stand, the dog is not barking in the night on this one.
ie yes gas prices are up, yes drilling activity is up, but new supply isn't coming on that stream that fast:
- there is a shortage of rigs, and people to drill with them - skilled manpower, geologists, everything else is in global short supply in an industry that hasn't hired in 10 years
- crystal meths has become a serious obstacle - there was a piece in the Financial Times about this, the degree to which it has become a real problem in rig crews. One of the states (Texas?) allows spot testing, and 1 in 3 crew members tests positive. The other (Oklahoma?) does not
- there just isn't the gas out there, no matter how much we drill. All the Canadian producers at least are screaming about the difficulties in replenishing reserves. One reason I hold on to my Canadian Natural Resources and Encana stock (two of the largest NG reserve holders in N America)
- shut in gas in the Gulf, post Katrina - not sure how big a factor this is
Posted by: John at May 22, 2006 02:50 PM
The casual observations made in the blog seem worth noting, but isn't it also true that these product markets are very much different (which seems to drive the divergence)?
The fact that the crude oil market is a global market, albeit with basis and quality differences baked in, and natural gas is a local market (due to lack of, and cost of, transportation infrustructure), would seem to explain what may just be a temporary difference. Hence, we still see $10 natural gas flared in some parts of the world!
Posted by: Bob Ames at May 22, 2006 08:53 PM
A fine graph, thanks! But it shows that the present discrepancy is not a new phenomenon. Gas prices have several times been about 50% lower than oil price/6.
Oil is clearly the main price setter here because oil is traded globally but gas not quite yet. So we should expect a rise in the US natural gas price.
There is another interesting point here: it is the industry that gives in, not the households, when energy beacmes pricey. Gas fired power plants are idle, but also so are fertilizer plants and other industries that use a lot of gas. Industry can close doors and move out of country, the households cannot. Energy-intensive industry is very energy-sensitive. Households are not because energy is just a small share of their expenditures. This has been a trend since the '70s.
Industry is also a trend-setter here through the investment decisions. The decisions NOT to invest because of high energy prices or uncertain supply have a long range impact. Therefore it might happen that we don't see the expected natural gas crisis (due to depletion) but its impacts will still be felt.
Posted by: TI at May 23, 2006 02:28 AM
The prices for oil and gas are set in the futures market, which is out of control. It is hard to extrapolate any sort of logical reasoning based on the physical market characteristics to the recent "market prices".
The "hot money" crowd has been allowed to dominate the futures market to the point where price is set in a non-competitive manner. Look at the obsene mess that has been underway at the LME and Comex since mid-march. I think the price peaks and vallies for natural gas suggest a market that is periodically raided by this crowd.
Thanks for the blog.
Posted by: ziggy at May 23, 2006 05:06 AM
I think you confuse cause and effect when you talk about the hot money crowd dominating futures markets.
One predicted characteristic of peak oil and gas production is increased volatility. Higher volatility is what ATTRACTS the hot money crowd.
Ergo, declining production rates cause demand mismatch and swings in price, supply, and demand. But that's a long story covered elsewhere.
Posted by: Joseph Somsel at May 23, 2006 09:16 AM
Could there be an oil benchmark issue too?
Are we assuming the sweets prime WTI and Brent prices instead of prices marked to the more sour cousins Urals and Mars blends?
With a refinery shortage, WTI ans Brent are at a premium b/c of their higher gasoline yield; by contrast the differential in prices to the sour grades has been increasing.
I am aware that EPA has maximum sulfur limits in oil for these plants; but, It's positively less than gasoline.
BTW, I know our city's power plant is in the process of converting, 15 % I understand, to gas from oil.
And, I agree last winter's mildness left a glut of gas that has to burn before we see any gains in gas prices.
Posted by: Joe Rotger at May 26, 2006 06:44 AM
Most "experts" believe domestic onshore gas production is up 3-4% year-on-year. Offshore is in something of a decline. But production is at least flat overall, and inspite of the hurricanes storage is as full as it has ever been.
We are looking at something of a Nat gas glut toward the end of the summer. At injection rates, some people project all available storage to be full in Sep 06.
Weather is the primary uncertainty.
Posted by: dis at May 26, 2006 12:54 PM
Good work and kudos to T.R. for digging up the Raymond James info.
I have posted my thoughts at the link below and cross referenced back to your post.
The Nattering Naybob
Posted by: The Nattering Naybob at May 27, 2006 10:45 AM
It's a consiracy. Based on your graph Ngas has increase with Crude Oil almost equivalintly. Yes there is a dope in the recent market and the interesting massive spike in '01 and'03. Ngas and crude oil are no where the same, two completly different sources and two completly different ways the earth produced the product and secondly crude oil must be refined to get anything useful out of it. So in short there should be no relationship with the two, just because crude oil goes up does not mean that Ngas price should increase.
Posted by: Layman at June 14, 2006 07:06 AM