This Week in Petroleum 11-21-07

Updated: Well, we got that big surprise, primarily because crude imports were sharply down from last week. Some excerpts:

U.S. crude oil refinery inputs averaged 14.9 million barrels per day during the week ending November 16, down 151,000 barrels per day from the previous week's average. Refineries operated at 87.0 percent of their operable capacity last week.

U.S. crude oil imports averaged over 9.8 million barrels per day last week, down 667,000 barrels per day from the previous week. U.S. commercial crude oil inventories (excluding those in the Strategic Petroleum Reserve) dropped by 1.1 million barrels compared to the previous week. At 313.6 million barrels, U.S. crude oil inventories are in the upper half of the average range for this time of year.

Total motor gasoline inventories increased by 0.2 million barrels last week, and are below the lower end of the average range. Distillate fuel inventories decreased by 2.4 million barrels, but are in the middle of the average range for this time of year. Total commercial petroleum inventories decreased by 6.9 million barrels last week, and are in the upper half of the average range for this time of year.

Even if we don't pop $100 today, at this level it won't take much volatility to take it above $100. At this point, I wish we would get it over with, so I will stop waking up at night and checking prices.


As oil stands again at the cusp of $100, here is what analysts expect for this week's report:

Analysts surveyed by Dow Jones Newswires, on average, predict that crude oil inventories rose by 800,000 barrels last week, while refinery use grew by 0.4 percentage point to 88.1 percent of capacity.

Gasoline inventories likely grew by 700,000 barrels, the analysts predicted, while inventories of distillates, which include heating oil and diesel fuel, fell by 400,000 barrels.

While oil supplies likely rose last week, prices were being supported Tuesday by concerns there would be a bullish surprise in the EIA report, such as an unexpected decline in inventories.

If we see that unexpected decline, then WTI should break $100.

Last week, I noted that even though there was a very big surprise with respect to crude inventories, the market seemed slow to react. I indicated that it may have just been an artifact, and that was in fact what it turned out to be. I get my quotes from the NYMEX site, and those quotes are delayed by 15 minutes. So, no opportunities to make money as a result of a slow-moving market. Actually, I would have been stunned if traders weren't poised to react to a large surprise in the report, but it seemed as if they weren't. That's why I posed the question.

In fact, Jaymax posted a very interesting graph that suggests that in fact the movement in price happened prior to the release of the report:

December WTI Following Last Week's TWIP Release

If that graph is hard to read, here is the link for the original graphic, in case you want to see the fine details. What it looks like is that about 4 minutes prior to the release of the inventory report, the price rapidly dropped over $1/bbl, implying that contracts were being dumped. This could of course be innocent; someone could have rolled the dice and guessed that the report would be bearish. That could also be due to the resolution on that graph (i.e., what you see above may have actually happened just after the report's release).

But for a suspicious person like me, I started wondering about just how many people have access to this data. It would be very lucrative to sell some advance information, so I am curious as to how the EIA safeguards the early release of the numbers. How many people know the numbers before they are released? What safeguards exist to prevent someone from selling the information? Do any of the EIA's employees drive a Ferrari? (kidding)

I asked Doug MacIntyre, author of This Week in Petroleum, if he could comment on this. Doug wrote:


EIA understands completely the seriousness of our data and carefully safeguard it before it gets released. We know that a lot of money can be made if the data were known prematurely, and everyone involved is very careful not to divulge ANY information to ANYONE before the release. In fact, we even go a little farther and try not to comment on the data to the press until at least 1 hour after the data are released. I am confident that the data were not, and have not been compromised.

EIA will not discuss the specific procedures we do to safeguard the data or divulge the number of people that have access to the data, as we believe that any information regarding the procedures we follow should be safeguarded as much as the data.

Thanks for explaining that, Doug.

A closed security system is more likely to have exploitable flaws, since fewer people have reviewed the security procedures. Flaws are hard to find, but easily exploited once found.

We could have no better example of this in the UK today:

A department of the government managed to lose the personal data + bank details of approx 25000000 UK people in the post!

If you want to get down to it, the reason they were sending all this data in the first place is because a manager decided it was 'too much bother' to send only the subset of data originally requested. Why was it too much bother? Because it was an outsourced contract and it wasn't covered by the usual terms, so it would need a special contract with the 'supplier'.

Those with a little IT knowledge will know that a SQL query to extract just the required fields would take 5 mins tops - but because we have a culture of ignorance in management and pointless outsourcing a major security risk is the result.

The relationship to peak oil?

You wonder why people don't question reserve numbers, don't think, don't make sensible plans? Its because everyone is busy following rote operations, trying to extract 'efficiency' out of old approaches rather than making new ways of working. They are doing it because an MBA teaches you to reapply and not question, to mechanically regurgitate rather than innovate. Nothing in an MBA, or the operation of most firms, rewards being smart - only being conservative in approach. Sticking you head above the parapet and saying 'but' is frowned on.

Make no mistake, we do it to ourselves.

Dogmatic thinking in a stable environment works fine. The same dogmatic thinking in a changed environment can be a big mistake. In which situation are we currently residing?

I would argue we haven't been in a stable environment in any industry for a few decades now. Too much change, new opportunities, new technologies etc. for that old paradigm to work. Yet they keep churning out MBAs with the same old failed mindset and tools. Its endemic, plodder begets plodder and civilisation ossifies.

The solution is very simple. Bet against the system by buying oil futures. If 1/4 of the population did this, oil would be too expensive to afford and we would be forced to adjust.

perhaps the overleveraged consumer can get a large cash advance on the nearly maxxed out credit card to accomplish this.

It's not just a problem at the post office, government, and business, but you may as well add academia to the list.

Never mind the details, suffice it to say that in my department, my mention of peak oil was scoffed at a year ago. Even lately, as I've mentioned again, the response is blank stares.

Smart people look the same as deer in the headlights given information outside their 'domain' of expertise.

Lovelock's analogy comes to mind...

As we weave our way through the tourists heading up to the castle, it's easy to look at them and feel sadness. It?s harder to look at them and feel hopeful. But when I say this to Lovelock, he argues that the human race has gone through many bottlenecks before --and perhaps we're the better for it. Then he tells me the story of an airplane crash years ago at Manchester Airport. "A fuel tank caught fire during takeoff," Lovelock says. "There was plenty of time for everybody to get out, but many of the passengers wouldn't move. They just stayed there in their seats as they were told to, and the people who escaped had to climb over them to get out. It was perfectly obvious how to get out, but they wouldn't move. They died from the smoke or burned to death. And an awful lot of people, I'm sad to say, are like that. And that's what will happen this time, except on a much vaster scale."

Never mind the details, suffice it to say that in my department, my mention of peak oil was scoffed at a year ago. Even lately, as I've mentioned again, the response is blank stares.

I actually had a real success yesterday. I took one of my former colleagues out to lunch and we had a nice chit chat about this and that. We talked about blogging and I showed him my energy blog and asked him if he had ever heard of peak oil. Vaguely he said he had. I then gave him a whirlwind tour of the topic in 30 minutes. I capped it off with the export land model as I understand it and before I was 2 minutes into the explanation he interrupted and said: "... that means that the declines for us would be experienced faster than a geological peak." He completely understood.

I showed him TOD and he started clicking through Stuart's or Khebab's recent posts about production ... intensely looking at the graphs. I mentioned ELP, he got it too. He said "Send me the TOD link."

I also mentioned doomer pr0n and cornucopians ... "where do you sit?" I told him on the fence!

What does your former colleague do for a living?

There was plenty of time for everybody to get out, but many of the passengers wouldn't move.

Lovelock often makes a creative interpretation of the facts to support his theories, but this time what he says is a mockery of the truth.

The fuel-fed fire rapidly developed to engulf the left wing by the time the plane had come to a stop, and smoke quickly spread into the passenger compartment. There were problems opening exit hatches, during the delay many passengers were overcome by smoke. Other passengers pushing to get out blocked aisles and exit ways. In later years, some survivors admitted they pushed passengers out of the way, climbing over them in a desperate scramble to escape. With a cabin filled with dense toxic smoke, it is certainly not obvious how to get out.

The passengers were only told to remain seated while the place was still moving, and before the crew realised the extent of the fire and the need to evacuate. As soon as it came to a stop, they were told to evacuate.

In short, there was very little time to get out, and if anything there was an unruly scramble. The idea that passengers sat passively as they were told and therefore died is complete rubbish. This is the complete opposite of what Lovelock claims.


There are a number of people who will know the data and a number of ways of secreting that data out, even if it is just a thumbs up or thumbs down.

There are a number of people who will know the data and a number of ways of secreting that data out, even if it is just a thumbs up or thumbs down.

True. But then it would be pretty easy to catch who is doing it: just follow the money. See who was selling all those contracts four minutes before release, and find out who their friends inside the organization are.

The fact that it was only four minutes before actually suggest to me that someone has figured out some technical glitch. Say the PDF is put on the server first, then it is linked to from the main page. While most people is waiting for the link, someone might be able to guess the URL and get directly to the report a fraction of time before everyone else.

Another possibility would be itchy fingers: traders not bothering to read the report but looking at the price to react quickly to the reaction. A small drop down and the trader sells, causing more drop and more twitchy traders selling and a small avalanche. 50% chance of being right...

If there is advanced knowledge of the report, and if preceding crude price changes are an indicator, this could be a lousy report.

Guess we'll find out, but I have suspected that there is advanced knowledge due to changes in gas station retails the night before the report is released in the direction the report indicates.

My two cents....

OK, now the report is out. Does this idea hold weight here? Can you watch the price of crude prior to the release of the report and guess at what the report will say?

If so, what are the implications? You brought it up Robert. Give us some thought here.

Hey Robert it's me again..I run OilGasFutures.Com, we're the commodity brokerage firm specializing in energy...a few comments on the inventory report and the market in general: Inventory reports will affect the market on the day they come out etc. however, as you and your readers know Oil is now a global market, with supply/demand dynamics continuing to override short-term market pullbacks (on inventory reports/ other news etc.) as as we go forward. The market continues to aggressively recover from pullbacks.

Many are asking why this time for $100 oil? How come a weaker US economy isn't curbing demand and lowering prices this time? The answer is that Oil has become more of a global market than in the past. The astounding growth of China and India coupled with a weaker dollar will keep prices high. Expect $100 per barrel and $5.00 per gallon at the pump to be the new reality for Americans and look for the market to test $120 to $150 per barrel soon.

There will be pullbacks along the way, which are just opportunities to get in or to add postions at a cheaper price. Crude is going up. pk

In the interest of remaining consistent, I will pose the same question I asked last week in a different thread.

If speculation is not the main reason for oil prices being way out of whack with the fundamentals, then please enlighten me and everyone else as to the real reason. The price of a barrel of oil should be around $35 or $40 based on the fundamentals and the costs involved in producing and shipping the product.

Apparently, large investors and hedge funds have allowed greed to take over and the common citizen is paying the price. I'm just waiting for the market to tank so I can watch the greedy SOB's cry like little girls when their investments go bad. Make no mistake, the oil futures market is in a huge bubble, and when it bursts like the sub-prime mess, I will take great joy while the greedy are punished.

Price has little to do with the cost to make it. It has to do with supply and demand.

See Beach Boy's post on Drumbeat for some good reasons why the price is where it is.

What "fundamentals" price oil at $35 to $40 a barrel?

Currently I believe there is more people wanting to buy oil then there is oil on the market.


Off Grid, Off Mainland, current profession:Beach Bum

I am increasingly of the view that the falling dollar may be the biggest driver of oil prices at this point, even more so than speculation and supply/demand tightness. The two have been moving conversely in lock-step for quite some time now, while there has been very little in the way of unusual speculation or untoward oil supply events.

I don't understand this point of view. Oil has gone from about $11/b to $97/b in the last 10 years. How does a declining dollar account for an 881% rise in the price of oil? And what is the dollar declining against? Gold? Gold went from $250 to $800, about 320%. If you are gauging the dollar decline against the Euro? Well that is about 150%. Are you talking about the last few weeks? And what about all the countries that have been priced out of the age of oil; Bangledesh, Zimbabwe, et al? That is real demand destruction, and yet the price is still going up.

I don't believe that a declining dollar is the cause of oil price rises. I think the continuing shortfalls in oil production makes it such that only those nations that can afford $100 dollar oil or more will have enough oil. But the world is no longer getting smaller. It is getting larger, and the cost of moving goods around the world is going up. I think the declining dollar is an effect of oil shortages, even if the US isn't feeling those shortages directly; others are. If the dollar value and the price of oil seem to be moving in lockstep for the moment, well OK, but is that really relevant? The value of the dollar is a matter of faith alone. Is that true of oil?

You say that the value of the dollar is driving the price of oil. I think you are putting the cart in front of the horse.

Sorry, I should have been clearer. Yes supply & demand fundamentals are mainly responsible over the last 10 years. I mean since roughly September 11 this year, when oil began its sharp up-spike. Compare:

US $:

Schaeffer Research had a decent short article about this (source of the latter graph above).

The concept is simple: as oil is primarily traded in dollars, as the value of the dollar falls, the price of oil in dollars must rise in order to keep the net revenue flat. An article by Gwynne Dyer in today's Peak Oil News put it nicely:

The recent surge in the oil price, which may see it reach $100 a barrel in the near future, is largely a mirage caused by the collapse in the value of the U.S. dollar. (The price of oil is generally quoted in U.S. dollars, but cost of a barrel of oil in euros or yen has risen far less dramatically this year.) But the longer term trend, which saw the price rise fivefold between 1999 and 2005, was driven by the tightening supply situation as demand raced ahead while production did not.

Energy consultant, writer, blogger

Hi Chris,

Certainly looking at your graphs they do seem to be trending in opposite directions. And it makes sense that more "less valuable" dollars would be charged for oil. However, the slope of the oil price increase is much greater than the slope of the dollar slide. So I would still argue that while the dollar slide would certainly have an effect on the price of oil, there is something more driving the price of oil.

Thanks, Ben

Ben, you're right, there is something more: speculation and fear and supply-demand tightness and all the other usual factors that don't affect dollar valuation nearly the same way. I didn't say that dollar valuation was the only effect, I said it was the predominant effect. But it's very hard to tease out one factor from another.

Nope. The price of oil is rising independent of the dollar. ...

Everyone at this site ought to read a few of the articles here...

The 'fundamentals' are really basic: world supply has not increased since 2005 and demand has continued to rise, primarily in China and India....

Is that news to anyone here?

>"The price of oil is rising independent of the dollar"

stiv, how can you separate the price of oil from the value of the dollar in order to even make such a statement?

If you compare the two charts I posted above, you'll see that the direction of both lines changed sharply together, and this happened in the absence of events to explain it otherwise.

The common citizen will eventually pay the price, but pump prices have not risen proportionally with the recent runup in crude prices. I know I'll be drawn into at least one conversation on the subject of gas prices over the holiday weekend with friends and relatives.

"Hey, oil is $100 a barrel but gas prices haven't gone up that much. You can't tell me oil companies and the government don't manipulate pump prices."

Any retort to that without invoking contango, crack spreads or refinery utilization? Something understandable to doctors, truck drivers and burger flippers alike? It is hard to convince people that what they see with their own two eyes isn't true.

... and I'm sure when and if you consistently ask the same question again next week, you will consistently get the same answer you got both last time and this time.

[EDIT] With apologies for overly grumpy reply
Jaymax (cornucomer-doomopian)

The price of a barrel of oil should be around $35 or $40 based on the fundamentals and the costs involved in producing and shipping the product.

A few points:

* While some oil indeed can be produced and transported at that price, other oil costs more (I believe I recently read that Canadian tar-sands oil is more like $60). By claiming that $40 is the price we should be seeing, you would also be claiming that we have no current need to produce that Canadian oil.

* While there is current oil that can be produced that cheaply, the oil to replace it costs more than that. One can argue that the fundamental cost of oil would be the greater of the cost of current production and the cost of replacement, unless you think there is no need to replace the current production.

* In a market with abundant supply of a commodity, prices tend to fall to the cost of production, and the low-cost producers stay in business and the higher-cost producers close the mine/cap the well/let the fields go fallow. The difference between the cost of production and the value to the consumer goes to the consumer. In a market with inadequate supply of a commodity, the price tends to rise to its value to the consumer, and consumers who derive less value from the commodity are priced out of the market. At least some of the difference between the cost of production and value of the consumer is transferred to the producer.

* We are transitioning from abundant to inadequate supplies of oil/liquids. The fundamental in question is no longer the cost of production/transportation, it is the value to the consumer. And oil is terrific stuff, of very high value.

LOL - you wrote your post while I wrote mine below - good to know that we arrived at same conclusion independently...;)

The price of a barrel of oil should be around $35 or $40 based on the fundamentals and the costs involved in producing and shipping the product.

That may be the AVERAGE cost for oil, because some of the reserves we are pumping were found long ago at $5 per barrel. But the MARGINAL costs are much much higher, and increasing. Look at the finding costs from 2005 to 2006:

In order to make a profit if the finding costs alone are $33, oil priced at $50 is not enough. Thats why if we go into a steep recession, oil company stocks are going to get creamed along with the rest of stocks -their revenues will have dropped while costs still extremely high. (maybe not so for nat gas co's)

I assume you are saying that price will fall if we go into recession, and this price decline will be tough on many oil companies, depending of course on their particular cost structure. The best guide we have for the near future is the seventies, where price went up 9x as the economy passed in and out of three recessions, generally driven by high energy price/reduced supplies. Price never declined y/y... what basis, then, do you have for suggesting that this time, as we pass po, a recession not even primarily caused by high oil price will manage to significantly depress price?

Further, the us uses a smaller proportion of world oil today than in the previous epoch, so demand in other growing economies, eg exporters plus chindia, is of increasing importance. Consider that recession in the us will not necessarily push china into recession as it transits from export driven to an economy that transfomrs a greater number of their citizens into modern consumers.

Consider possible numbers: if the us consumption declines 4% on account of recession then world demand declines 1%, or approximaely what we have been seeing in c+c since mid 2005. Then, possible further decline plus declining exports... hard to see prices going down.

Summary of Weekly Petroleum Data for the Week Ending November 16, 2007

Crude: -1.1m

Gasoline: +200k

Distillates: -2.4m

With the Thanksgiving weekend around the corner, next week's inventory number could show a brutal decline in gasoline.

6.9 million barrel decline in commercial inventories - tiny increase in gasoline, 1.1 mill decline in crude, BIG aggregate decline in distillates, aviation and propane. Ouch. Fundamentals in place for a push over $100 either today, or next week.

$99.59 so far... about $1.20 jump

The price is dropping on this report? can anyn explain that .

30 minute delay. Just wait a few minutes.

what price you looking at? wti spot was $99.59 @ 10:50 according to upstream. The charts on the right are front month (jan 08)

API and Cushing showed a build in crude (per cnbc). So there's a possible discrepancy in numbers.

API shows 1.4 million build.

I was looking on the right, but now I see it has quite a delay. sorry

Oil wavers after surprise inventory drop

Energy futures wavered, hesitating on a drive to $100 a barrel Wednesday after the government reported that oil inventories fell unexpectedly last week, but that supplies at a closely watched oil terminal in the Midwest rose for the first time in weeks.

"EIA will not discuss the specific procedures we do to safeguard the data or divulge the number of people that have access to the data, as we believe that any information regarding the procedures we follow should be safeguarded as much as the data."

So much for governmental transparency...

SubKommander Dred

the other report---the API report showed an inventory build....see my thoughts above for market/inventory


What's up with the AP release?

In its weekly inventory report, the Energy Department's Energy Information Administration said oil inventories rose by 1.1 million barrels during the week ended Nov. 16.

Is this shitty reporting or intentional?

Bloomberg is carrying a story saying essentially the same thing, stating that there was an inventory increase at Cushing, Oklahoma. Then, in the second paragraph of the report, goes on to directly contradict this, stating that overall there was a stock draw down of about 1 million barrels.

SubKommander Dred

Yes- I was frustrated because posters at TOD referred to the 1.1 million crude inventory drop but both MSNBC and CNN.Money were reporting a 1.1 million gain. The headlines at both sites contradicted the text of the news and interestingly both read almost word-for-word the same.

Shoddy reporting?

It's the last real trading day before Thanksgiving, which is followed by Black Friday, the biggest retail shopping day of the year. Look at the retail stocks. They are sucking wind. If you were the Bush Administration, would you want $100 oil to be the topic of conversation on the biggest shopping weekend of the year? So you sell a few contracts to cap the price. Happens all the time.

RR - your (valid) point about data security notwithstanding, I don't think that chart shows any reason to worry - eyeballing it, the relevant interval covers 46 minutes, and it looks like the first big slide happened about 10% into that interval: 10:26 + 4.6 minutes = 10:30 near enough.

Each tick on that chart is two minutes. There was a significant drop at 10:28.

Jaymax (cornucomer-doomopian)

The chart not withstanding, a drop was not observed in the real time board reporting. There were a couple of quick spikes (down and back up) that lasted about 1 second. But most od the variation I saw was around $0.05.

Using the countdown clock on CNBC (and updating the EIA site) I watched and recorded the price at 1:00 minute intervals all the way to 0:00. The lowest price was 97.32 (using this method of record) at 10:24 AM.

At 10:00 AM the price was 97.82, it dropped to 97.32 and then rose slightly to 97.47 before dropping to $97.37 in the last minute. There was an immediate spike to 98.70 before dropping back.

So, I did not see what is described here watching it live.

The drop at 10:28 was last week, not this week, when it spiked DOWN to about 92.40ish.

Jaymax (cornucomer-doomopian)

Anyone want to go in on a bunch of oil futures so we can force Robert to pay up before thanksgiving :)

We should have enough readership to sway the market a few cents :)

Well dang not gonna make it :)

Robert enjoy your Thanksgiving in Scotland. Hopefully you have enough expats around to get a holiday spirit.

For comparison, here's today's.

Because it's sooner after the event, I haven't had to squish it so much, and each tick is 1min rather than 2min.

What's perhaps most interesting, is again there's a significant jump just before the report comes out (note that the open and close marks are right at the top and bottom of the 10:29 bar), but this time it goes in the wrong direction.

I wonder if the inherent pent-up volatility in the market at the time the report comes out, combined with everyone expecting to be moving in the same direction as it were, can leads to a significant shift in a random direction in the minute or so before release.

Jaymax (cornucomer-doomopian)

Yes, I was watching on CNBC and they went to the high-speed sweep of prices on the screen (where the info on the screen represents about 5 seconds of data.

But the spike (in less than 5 seconds) was very, very impressive.

If I was doing this at a large institution, I'd have a program running probing the EIA website as often as they'd let me, and ready to parse the scraped HTML report, and feed the numbers right into my trading algorithms. Done right, the largest delay between when the numbers showed up on a web site in Washington DC, and I made my first trade based on them, would be speed of light in fiber between DC and NYC (something on the order of 10ms). If the EIA rate limits how often I could connect to their site, it might well be worth it to have connections coming from a large number of IPs so they couldn't tell so easily.

Thank you for that free consultation :)

You have a lot of faith on network and computer speeds and latencies! You have at least three machines and two long internet connections in the way. I would say the EIA web server plus whatever system the trader uses to record transactions will delay you much more than light travel time. And yes, doing a query for a non-existant page every 1ms should trip some alarms somewhere.

If you got the whole thing under 100ms I would be impressed.

(I happen to know a lot about latency in distributed applications, since I started a small cottage industry in estimating how fast computer worms can spread with this paper, and I've also worked for companies that design network equipment so I have some feel for device latencies).

There's nothing to be done about the EIA webserver except probe it as often as it will let me get away with. Presumably, over time the EIA has sized their system to whatever traders do. There's also nothing to be done about the exchange computers, but my mental model of that is that whoever consistently gets their request in the queue first at the exchange computers will make the most money. So the key to making money at this game must be to get my overall time as close as posible to the latency between the EIA website (I assume in DC, but if it's somewhere else it doesn't change the basic calculus), and my systems in NY.

Presumably my systems are very close to the exchange, so the second hop is not a big factor. You might be correct that traders have databases (which go to disk) in the loop. If so, they are throwing away money. A well crafted custom application that doesn't go to disk should add less than 1ms to the overally latency - I control my own systems so I can make sure it's lightly loaded. If they aren't doing it that way, I would speculate that money is being thrown away. It would seem to me that if my competition consistently takes 100ms to get an order to the trading exchange computer, and I could cut it to 20ms (or whatever the exact calculation turns out to be), I ought to make a lot of money, right?

They are onto you. They block ICMP traffic, but from me it's a least 70ms (at least 19 hops) and I have a gig connection. Interestingly, it's only 22ms to NASA from me (12 hops).

You don't need ICMP - just lots of HTTP requests in flight constantly, so when the change in the page finally comes, the time between then and your next HTTP request for it hitting the server is as small as possible. Time of flight for packets is much longer than the time between successive issue of packets for fast applications, so the goal is to keep the pipe between you and the server as full as requests as can be got away with.

Also, hops are irrelevant, since core routers add negligible latency (sub 20 microseconds typically) - long haul latencies are dominated by speed of light (unless there is congestion, but we might assume that big firms can buy ample capacity)..

70ms sounds like you are on the west coast.

70ms sounds like you are on the west coast.

Yes I am so it's not good for me. Leaving aside the in between routes, I wonder how the LANs play into it? Maybe for such a little page it wouldn't matter.

Good idea Stuart. Maybe I should get some of my staff to write something! -;)

Being on the west coast is bad, because the information first has to cross the continent to get to you, and then cross back again for you to execute a trade in New York. So someone playing the same game in New York has you beat cold.

However, all you would have to do to solve that problem is lease suitable colo space in NYC. More serious is making sure you have excellent links to where a trade is going to be executed (any chain of middlemen will kill it also).

More seriously, I would be very surprised if the big players haven't been all over this kind of thinking for years.

They have this piece of script that counts down the seconds (and then auto-refreshes a typical browser) the 'moment' the report is released.

By working out nominal latencies and their system clock offset, and testing often enough to establish their script logic and timer consistency, you shouldn't need to poll at regular fast intervals, you could possibly tune the system to ensure your well timed request hits their server the moment the countdown script expires.

I guess you preconfigure in the 'market expectations' do a fast weighted judgement around which numbers are above and which are below and buy or sell an amount derived using an algorithm based on the delta against the expectation.

Of course, you shouldn't need to wait for the whole page to load - but probably worth avoiding the text scraping from the summary at the very top, most of the key numbers are repeated in well-formed tables about 2KB in...

Jaymax (cornucomer-doomopian)

Or...easier have a source in the EIA that lets you know the final figures the night before the release...then you get your fronted hedge fund to queue up the trades the millisecond after the report is released and there is no need to read the text.

Last time I saw him, Clarence Beeks was on the way to Africa so we can't use him to get the report early.

Even using a co-lo down the street from the exchange might not help as much as one would think. I'm just down the street from Georgia Tech... I can see the campus out my window but it takes nine hops to get from my computer to their web server. Back when I was on AT&T broadband it was even more hops. A trace route showed it following a path all the way up to Virginia and back to Georgia to move packets between computers that are only a couple thousand feet apart. The net can be a mysterious and convoluted place.

Someone who was trying to game the system by probing the EIA website could certainly try and buy access on a machine that is "close" (i.e. small number of hops over big pipes) to the EIA webserver. They could run the app there, and then send the results on once they were known.


I think I would just sleep with the girl that makes the copies :)

Gathering information that appears on the web can be done in lots of ways. People are usually looked to as a resource in inside info.

If you were a large institution, such millisecond accurancy would be irrelevant because you'd have 10,000 contracts to move. You might get off three or four.

A "market" is not some computer game. For every seller there's a buyer. Why would the person you're transacting with make the trade without reading the report for themselves? Because they want to give you their money?

Interesting point. And I should stress I know next to nothing about trading systems per se (except I once got a tour of the NYSE computing faciliities). However, looking at the graph up above that Jaymax posted, it's clear the price is moving right up to the point the report is released, so there's trading going on right through the report release. Which implies there are people (or systems) who have not read the report putting orders in just before the report is out. Presumably, those orders are some random mix of buys and sells. So, it still appears to me that trading with those people/systems once one has an additional delta of information that they don't have is a profit opportunity. Of course, once all those orders are used up, and there's only orders from people who already have the news, then the profit opportunity is gone.

I agree with you that there many orders to submit though - I imagine a whole raft of different stocks, currencies, etc all need to move a little bit because of this change in the price of oil, so there would be many things to do. My recollection from NYSE is off a massive hall of Unix servers doing nothing but take orders from all the institutions connected to them. So it's not obvious to me why one wouldn't be playing the same game in every case.

Well, the price today has tanked $2 on an inventory draw?!?. Anyone care to unravel this mystery? (likely the cushing build or profit taking before thanksgiving!)


The stock market is tanking, house prices falling, leading economic indicators falling. These are all bearish for oil prices. We'll need another chunk of bad news to push us over $100.

True - there are many other factors like currency fluctuations, weather, geopolitical, profit taking etc. There's also huge money and interests that don't want that $100 level breached and having it on the cover of Time, Newsweek, etc. and lengthy articles written on the subject for the masses.

One possibility.

Bush's PR guys didn't want Harry and Marge, and Pat and Jim, and Laweesa and Jermaine, and Tony and Marie, and Jose and Conswerta, and Tim and Tom, and Lisa and Laura talking about $100/barrel oil over slices of tryptophan and can-shaped blobs of cranberries.

Correct...a 200+ drop in the DOW is a bit more palatable. It appears these days, it is either one or the other. Either the DOW takes a big hit (which in turn brings down the price of crude slightly) or the DOW gains (which in turn pushes the crude price up close to $100). We are currently on a see saw that swings on a daily basis.

I am a firm believer that these related patterns are not random.

Stocks are down below 2005 levels but still around historic average levels:

However, domestic production has a big drop this week (-192 kbpd)!

and crude imports are still way below were they should be (i.e. around 11 mbpd):

I'm going to be really pissed when you guys start charging for this info Kebab.

Sorry...KHEBAB, your much too cherished for me to get your name wrong.

Would be nice graphs if not for the grey "urban digital pattern camo" behind the lines, which makes them really hard to read :{

1st post...

However, domestic production has a big drop this week (-192 kbpd)

I wouldn't make too much of this because lots of producers of less than 10bpd wells probably took advantage of the run up on Sept and Oct prices to sell a load of oil by topping off a load from one lease with a partial load from another.

I know we did.

Interesting 1st post.
Thanks and Welcome.

Watch that domestic production according to my model it will continue to tank.

I just read this on Bloomberg:

The Organization of Petroleum Exporting Countries will load 24.5 million barrels a day, compared with 23.8 million barrels in the month ended Nov. 10, Oil Movements said. It will be OPEC's 14th consecutive increase and the biggest this year, according to the company.

Could anybody explain this ? It seems to run counter to much of what has been reported on TOD recently. To what does the 24.5 mb/day refer ? Who is the company ?
It just doesn't make any sense.
Yours Puzzled.

Going by this graph from TWIP it would seem that we are now at the time of the year when Gasoline Stocks begin their largest build of the year. Currently stocks are about 195Mb. The peak occurs around Feb 1st, which is in 10 weeks. Then stocks typically decline until May. To achieve the highest average of around 220Mb, stocks need to build by 25Mb. This means an average build of 2.5Mb per week from now until then.
I see that from the graph a large build did occur during last winter, and that still resulted in a summer with gasoline averaging over $3.20 at its peak.
My question is to Robert or anyone: What sort of gasoline build do you envision over the next 10 weeks? And what number needs to be achieved to avoid more widespread shortages than occurred last summer. Will 220Mb even be enough?

US fuel inventory levels, measured in days, seem to have hit a plateau, similar to world total liquids production as forecast by al-Husseini and myself

The figure below shows the downward trend in days supply of gasoline since 1991. Since 2003, days supply have not gone below 20 days but since Feb 2007 the 52 week moving average has kept moving down. If days supply drops below 20 days, this downwards breach of the historic short term minimum operating level could cause gasoline shortages. These shortages would not necessarily occur nationally but would probably occur in states where distribution bottlenecks exist. This already happened in North Dakota in July this year

(A description of this Minimum Operating level clipped from an EIA publication follows. “…maintaining minimum operating levels (e.g., gasoline must be present in the pipeline at all times to push product further through the pipe.) When actual inventories drop below minimum operating levels, the system effectively may be running on empty. EIA reported that PADD II inventory levels in May and June 2000 were at or near minimum operating levels )

click to enlarge

The next figure shows a similar downward trend in days supply of distillates.

click to enlarge

The last figure shows an average, for illustrative purposes only, of the first two figures.

click to enlarge

All the figures above show that there may be a very small excess capacity of fuel available in the USA. If any shocks to the distribution system occur, then fuel shortages will probably occur in some states.

If domestic gasoline demand actually responds to the rising oil price, what does that mean for refiners? I would think we would find that we have surplus refining capacity, and crack spreads would become very poor. Perhaps we should be shorting refiners?

No. We shouldn't be shorting ANYTHING. There shouldn't be any money going into the NYMEX or oil stocks, period. If everyone pulled their money out of these markets, the currently fictitious price of a bbl of oil would correct itself to $35 or $40/bbl where it SHOULD be. The NYMEX is full of greedy people, and I for one am sick of seeing wealthy investors manipulating energy costs at the expense of the common citizen. It's time to stop playing games for the sake of greed. Doesn't anyone see what is REALLY going on here, and doesn't ANYONE give a crap?

No. We shouldn't be shorting ANYTHING. There shouldn't be any money going into the NYMEX or oil stocks, period. If everyone pulled their money out of these markets, the currently fictitious price of a bbl of oil would correct itself to $35 or $40/bbl where it SHOULD be. The NYMEX is full of greedy people, and I for one am sick of seeing wealthy investors manipulating energy costs at the expense of the common citizen. It's time to stop playing games for the sake of greed. Doesn't anyone see what is REALLY going on here, and doesn't ANYONE give a crap?

First of all... no offense but it might be a good idea to consider reading some of many informative articles on this site. Unless you work as a journalist for CNN or Fox (which requires ignorance), it is clearly erroneous to believe it is merely speculation driving up the price of oil.

Here is a basic principle of markets. In order for a contract to exist, such as an oil contract on the NYMEX exchange (which doesn't "contain" people as you alluded to), there must be an agreement between buyer and seller to deliver a defined product at a defined time. The price at which this contract is agreed upon is the equilibrium price. Both the seller and the buyer of the contract believe that they are getting a good or at least fair deal, otherwise they would not enter into the contract.

I'm just wondering who it is that you consider to be greedy? Is it the people selling contracts or those entering into an agreement to buy contracts. The fact of the matter is that oil futures trading is a zero sum game. When you and I enter into a contract, either I profit because I made a correct prediction of price movement or you profit for the same reason - we cannot both win. If the common consensus was that the sellers of oil contracts were being greedy and trying to receive a higher than market pirce, other sellers would step in and sell oil at a lower price, and you might witness lower prices. The fact that this is not happening is not due to greed but due to pure market fundamentals supported by dimishing oil supply and a sharp increase in global demand. (not considering a weak dollar which most oil contracts are denominated in. Also, consider the fact that one barrel of oil contains the same energy of 12 men working full-time for one year, and suddenly $100 becomes absurdly cheap, let alone $35).

Consider this definition of greed; "excessive desire to acquire or possess more (especially more material wealth) than one needs or deserves". Based on what I explained above, does it not seem illogical and impossible in a free market that high oil prices are the result of greedy investors?

No offense,
But surely if more people are investing in commodities like oil, over stocks, there will be an increase in price.

No offense, but how do you think the large oil companies will do with oil at $35? (they are treading water with oil at $99).

Treading water ?? I think record profits is what you meant to say !

Were people investing in actual oil (ie: buying oil, putting it into storage tanks, this would be true), but they're not, and you're just simply wrong - that's just not how the market works.

For stocks, as for actual physical oil, there is a limited amount available, so the more investors, the higher the price goes. For futures contracts, there is no limit whatsoever on the number available - and so the number of investors in and of itself CAN NOT cause an increase in price.

No offence.

Jaymax (cornucomer-doomopian)

None taken , of course !! So you are saying that the number of investors; people willing to buy into, a limited supply of something, will not have any effect on the price ??

No, I'm saying exactly the opposite.

The number people willing to invest into an UNLIMITED supply of something, will not have any effect on the price.

I thought that was pretty clear - "there is no limit"

There are an UNLIMITED number of oil futures contracts available. Oil Futures Contracts ARE NOT Oil. It is a fallacy to suggest that because investors are buying into Oil Futures Contracts, they are buying into actual oil. They are not.

If you want to suggest that there are oil speculators buying up physical oil, then you are wrong - if that was the case there would be increasing oil inventories.
Jaymax (cornucomer-doomopian)

Hmm. There might be no theoretical limit to the amount of oil futures contracts available, but the amount actually supplied to the market is limited to the amount of money invested in the futures market, which is also still less than the total money in circulation.

The oil futures contacts would also be limited by the computer storage or filing system used to archive and access the information on contracts.

The oil futures contracts would also be limited by the time and resources available to the people who create, buy, and sell the contracts.

So, tell me again how the contracts available are "unlimited"?

So, tell me again how the contracts available are "unlimited"?

The practical limits you allude to are nowhere near being reached, therefore futures contracts remain fundamentally unlimited, as you have agreed.

As Bob points out, you appear to accept that any effective limits on available futures contracts are so abstract as to have no bearing on the price of those contracts. Which was the point.

Jaymax (cornucomer-doomopian)

No, the point was there is no limit, and limits do clearly exist.

So maybe what we mean to say is the effect of oil futures contracts on the price of oil is negligible?

Like a colony of ants trying to push a tricycle, it's having an effect, but it doesn't register in a significant way.

The point was that the number of investors [or total amount invested] in oil futures, of itself, has no effect on the cost of oil futures, because [effectively] there is no limit to the number of futures contracts available for sale.

Contrasted with any stock or physical item, where the number of investors [or total amount invested] by definition must increase the cost of the item.

Not negligible effect, no effect. The futures market itself ensures that is the case.

Note that all of this is about the cost of oil futures contracts - to what extent that drives or is driven by or otherwise is connected to the cost of actual oil is a separate point. But there I quite like your analogy of the ants and the tricycle.

The reverse works as well I expect, supply and demand for actual physical oil is like the child on the bike - any ants that stay on-board are going to end up where the child wants to go, and quite a few will get squished as a result!
Jaymax (cornucomer-doomopian)

Spin: Why not $15? The $35-40 range you are quoting was considered quite high until 2005. Don't worry- a giant CTL industry is coming on stream once prices hit $70 (and the USA has lots of coal). This is all good- the USA will become the KSA of coal.

Prices have already hit 70 dollars. KSA of coal !! I for one will be relieved when the good old US of A is no longer King of anything. How thin that veneer of morality is spread over the pile of greed and hypocrisy.

Sarcasm alert.

I'm disappointed that there is such a diverse and perverse spectrum of speculators and investors who are profiting by artificially inflating the cost of a barrel of oil. The only entities that should be profiting are the ones that spend the money and effort to get oil out of the ground and bring it to market. Instead, there is this global market that almost ensures that the wealthy get wealthier at the expense of the common citizen. In other words, the wider spectrum of humanity is getting screwed so that the few wealthy and privileged can profit off their backs. Doesn't anyone else see something wrong with this? Oil is not a discretionary commodity, it is a necessity. Therefore, it should be handled differently and should not be subject to the whims of the minority of wealth who can manipulate the price to line their pockets.

You ask: Doesn't anyone else see something wrong with this?

The thing you're wanting other people to be upset by doesn't actually exist - it's a convenient excuse for people who can't bear to face the thought that the price of oil might actually be fairly close to that determined by the market fundamentals of supply and demand.

You are outraged by your belief in a fiction.
Jaymax (cornucomer-doomopian)

I have to admit it seems oil prices are being driven by greed. Obviously there is also a supply problem; but it does seem like the recent run ups in price do not accurately reflect supply and demand. Do refiners buy crude on the Nymex exchange ?

Crude down 1.1 mb
Gasoline up 0.2 mb
Distillates down 2.4 mb
Propane/propylene down 0.4 mb

Total: decline of 3.7 mb
Commercial inventories declined 6.9 mb

Anybody know what happened to the other 3.2 mb?

All the data is in the last part of Table 1 of

EIA text report

or index to text, PDFs, tables etc here

Kerosene-Type Jet fuel down 2.1 mb
Residual Fuel oil down 0.2 mb
Unfinished oils up 1.0 mb
Other Oils down 1.8 mb

Jaymax (cornucomer-doomopian)