Oil output – your thoughts?

If you look down below, you’ll see one of my first second post, Statistical Battle I: Rig counts. By the title, you might guess that I’m planning a few other Statistical Battles. The next one will be on Venezuela’s oil output. And I need your help.

Anyone who follows Venezuela knows that the country runs on oil money. So everyone wants to know how much money that is. It’s not simple to figure out.

The Energy and Oil Ministry publishes a weekly price for the standardized “Petroleum Basket” that has little to do with the proportions of items actually exported. An oil ministry official told me that the makeup of the basket was published years ago in the Gaceta Oficial, but I have never been able to find it, and no one I’ve asked has had it. So the price per barrel has error bars of an indeterminate length.

But if you want to see error bars, you need to look at the oil production and export figures. Here are a few outside estimates, measured in barrels a day:

International Energy Agency (PDF) 2.23 million
U.S. Energy Information Administration
(only updated through December, excludes condensate)
2 million
OPEC (PDF) 2.29 million
Bloomberg 2.29 million
Platts 2.22 million
Petroleum Intelligence Weekly 2.65 million

The government has started sending (through an oddly back-channel method) a monthly report on how many barrels of each type of crude and condensate were reported for royalty purposes, the name and cargo of each tanker ship that embarks from a Venezuelan terminal, the domestic dispatches at each fuel terminal, and an independent tally of the bills of lading of outgoing and incoming tankers.

In the latest mailing, the government says oil production in February was:

Ministry of Oil and Energy
(and no, the link doesn’t go to the report)
2.89 million

2.89 million barrels a day — 9% higher than the highest of the outside estimates, and 45 percent higher than the estimate from Washington. That’s error bars. But it’s almost identical to the 2.87 million that you get when you add up the country’s self-reported net exports and domestic sales. So at the least, the ministry is being internally consistent.

When the government started sending its report, I Petroleum Intelligence Weekly revised its estimate upward by something like 500,000 barrels a day. If the others had followed, at least there would be a cluster vaguely near the official number. But instead, most of the outside estimators have glanced at the Venezuela government mailings and more or less dismissed them without public disclosure or discussion.

I’ve heard that back when the government claimed the production was 3.3 million barrels a day, a bunch of experts got together in Caracas to discuss the numbers, and concluded that the EIA-IEA-OPEC range was several hundred thousand barrels a day too low. As far as I know, no one has held a similar discussion since the government started releasing its monthly numbers.

Nobody has the whole picture. Given that I am seeing lots of big names in the access logs of this website, I hope that some smart people can write with their thoughts of what is going on. If you prefer to send an e-mail rather than leaving a comment, that’s great with me — settysoutham@gmail.com. Your confidentiality is assured if you need it. The only thing I don’t welcome is uninformed opinion. Be in touch.

17 thoughts on “Oil output – your thoughts?

  1. saurabh

    How remarkable is this? I’ve never tried to chase something like this down, but I have to imagine that coming up with a tally for any oil-producing nation is difficult and highly political. I doubt Aramco is any more forthcoming with their data…

    1. sapitosetty Post author

      Welcome Saurabh, wonderful to see you. Hey everyone, click his link.

      As far as I know, the monthly Venezuela data release is unique in OPEC. And yes, Saudi Aramco is also a black box. However:
      – Saudi self-reporting and outside estimates don’t show such a complete divorce from one another. As you’ll see when I post a future graphic, the estimates and the self-report aren’t just uncorrelated, they are quite reliably anti-correlated. And the trend lines run at completely different angles. I mean they have nothing to do with one another.
      – Other countries have more international oil companies in positions to know what’s really going on. After the 2002-03 oil strike, Venezuela consolidated control over its industry. At this point, there are very few international executives or laborers left in the country, cutting off a line of information for the international community. The government here is very sensitive about information, so Venezuelans are reasonably reluctant to speak: my slogan here is “those who know don’t speak, and those who speak don’t know.”
      – The number is politicized. The government here says the problem is that the estimators rely on a bunch of knucklehead Chavez-haters who want to sabotage the regime, cutting international confidence and bond ratings and encouraging coup-mongers within the country. The international estimators don’t talk much, but the official line is generally that governments lie, and the government here has long had an interest in showing that it recovered well from the oil strike of 2002-2003.

    2. sapitosetty Post author

      Oh and also – Wall Street cares more about Venezuela because it’s a volatile bond market. They love them some volatility.

      Oil companies care about it because they are thinking about sinking billions and billions of dollars here, and need to know what they are getting themselves into.

      Venezuelans care about it because it’s their livelihood.

  2. Jakob

    Setty, nice blog, interesting reads and highly informative. On the oil production stuff, this might be a good paper, looks like the authors did something pretty similar: http://www.cepr.net/documents/publications/venezuela_2008_11.pdf

    On the rig count post, I was looking at the OECD monthly oil market report (http://www.opec.org/opec_web/static_files_project/media/downloads/publications/MR042010.pdf), which includes a nice chart on rig counts. It uses the Baker Hughes numbers, I think…but I noticed that from February to March, while Venezuela’s count did go down, the world total also went down by the same percent …I’m no oil expert, but thought that might be of interest.


    1. sapitosetty Post author

      Hi, thanks Jakob. The CEPR paper isn’t up to snuff. The single biggest problem is that it ignores Venezuelan petroleum imports. It also uses as its source on oil exports to China a single article from the Xinhua newswire, which reported the number from a press conference here in Caracas. In a technique that wouldn’t stand up to peer review, they cite a Chinese Commerce Ministry statement even though it was actually a reported article reposted on the Commerce Ministry website. Chinese customs figures for that period showed much lower imports — lower by a good 200-250,000 barrels a day. There are other problems too, but these give the biggest room for error.

  3. Rebecca

    One of the big difficulties with tracking down oil production and export numbers, in my experience, is whether or not to count Aruba. Oil processed at the joint BP/PDVSA facility in Aruba is counted as Venezuelan output by Venezuela, but not anyone the US EIA. If you include it, it works out to Venezuela’s numbers.

  4. Jakob

    Thanks for the quick response, I’ll definitely defer to you on the paper. What about the rig count though? Doesn’t that seem to contradict the statement from your rig count post that “This month, the world outside the U.S. rose 3.3 percent, while Venezuela fell 21 percent.” I don’t know if one measure is better than the other, but at least there are conflicting reports?

    1. sapitosetty Post author

      I only showed the Schlumberger numbers originally because Baker Hughes measures the U.S. weekly on one sheet, Canada monthly on another, and the rest of the world monthly on a third, so it’s a bit of a hassle to graph it as nicely as the Schlumberger one.

      The point remains. Look at the graph of the Venezuela Baker Hughes count. It’s still way down from the historic trend, as if oil prices were low. The rest of the world is coming back, most notably the U.S., where use of oil-drilling rigs has doubled in a year. The the world outside the U.S. and Canada is also up 6.1% on the year and marginally on the month.

      Don’t worry too much about monthly noise. Rigs being moved, bad weather, etc can have a big effect. But the trend is clear on these charts — they say Venezuela isn’t mobilizing rigs like the rest of the world. The government, meanwhile, says this is a plot to destabilize the revolution.

  5. Rebecca

    I’m sorry, I meant Curacao when I said Aruba. And I should have been more specific. Last time I checked, the US EIA total for Venezuelan exports to OECD countries was the major source of difference between EIA and PDVSA numbers. US EIA doesn’t count the Dutch Antilles as part of the OECD, although they do count the US Virgin Islands. Since the Dutch Antilles are home to oil refineries, this makes a difference. I haven’t looked at this for a while, so this is more of a suggestion than an answer.

  6. sapitosetty Post author

    Hi Rebecca, I’m glad you could comment. I am not sure I understand what you are saying. Isla handles crude from all over, but of course primarily from Venezuela. Are you saying that PDVSA considers everything they process there to be Venezuelan production? I have a hard time believing it.

    On the sheets I’ve seen, I am pretty sure PDVSA considers shipments to and from the Antilles to be exports and imports. This is the main reason why I feel it was a failure of your paper to count exports without subtracting out imports.

    There’s also the issue, with China, that some of the PDVSA barrels to China don’t ever physically touch either Venezuela or China. They may be pumped from Ecuador under the Ecuador-Venezuela swap contract, handed to CNPC, and shipped to Singapore. They can still be counted as PDVSA barrels to China. This makes the sort of calculation you did (and which I did too before I understood these arcane aspects of the industry) much less useful, sadly.

    1. Rebecca

      My only involvement with that paper was to try to match up PDVSA’s country-by-country export data with the EIA’s numbers for OECD imports from Venezuela. What I found was that they matched, as long as I didn’t include Curacao in the OECD – but did include the US Virgin Islands (ie St Croix). This struck me as weird, so I asked someone at the EIA. Their response was that my calculations were right – they didn’t include the Dutch Antilles in the OECD, but they did include the US Virgin Islands. So basically I was able to verify that the PDVSA country-by-country export numbers matched the EIA numbers for OECD imports from Venezuela.

      Your other points about the paper sound reasonable, but they’ll have to be taken up with the paper’s main author.

  7. ow

    I don’t believe there is any controversy at all about how much money Venezuela makes from oil. They publish full financial statements each year that are fully audited by a subsidiary of KPMG.

    Given that you know how much money they make from exporting oil, which we do because their financial statements break down how much they get from exports and how much they get from overseas operations, to calculate how much they must be producing is fairly simple given that we know the price of oil. Sure, there is some room for error given that the price might not be exact, but the variance is unlikely to be more than 100 to 200k at absolute most. So if you work backwards from their financial statements I think you’ll see that their production numbers are pretty accurate.

    Further, they actually have their export shipments audited. This information is not from some “back channel” but rather by a British firm that specializes in these sorts of audits. Those numbers also confirm PDVSA’s production numbers.

    Finally, anyone who really thought PDVSAs numbers were bogus could just match their exports to each country (which their financial statements supply) to what each country has received. It is very easy to do that in the case of the US and they matched exactly. I can’t do similar matches that easily for other countries because that information is likely not on the Internet – but any organization that seriously doubted PDVSA’s numbers and wanted to show them to be wrong would probably have the resources on the ground to find out if, for example, PDVSA sent as much oil to England and Germany as it says it has. To date, no one has caught PDVSA in a lie via that method.

    There is a more to say about this, including how PDVSA’s oppenents in Venezuela have been busted in a significant number of lies, but that is it for now…

    1. sapitosetty Post author

      Hi OW. The financial statements are a fine place to start, and I often use them in the absence of anything better, but don’t trust those auditors. Enron, Lehman Bros., Stanford Bank and Bernie Madoff’s firm were all audited. When you are the biggest customer at an audit house, you get your way. The accountants are not a subsidiary of KPMG, but an independent affiliate. Without PDVSA, their income will shrivel.

      We have no idea how much money PDVSA gets from oil. All Caracas swap market participants I have spoken to agree that PDVSA must intervene in the parallel currency market, as there is no one else who needs to convert hundreds of millions of dollars into bolivars every year. Every time PDVSA does that, it gets bolivars for cheaper than the official rate, and is able to massage its books. That alone means there’s no way to backwards-engineer PDVSA’s finances, without even getting into the less provable guisos that the company may be involved in.

      They don’t have their export shipments audited. The reports from Inspectorate are very clear that they are not audits, but tallies of documents that are handed over by the oil ministry. I doubt that there is much fakery going on in that tally process, but there are definitely ways that the system could be gamed, most easily by not handing over some records of import shipments.

      PDVSA’s never been busted in a lie via that method? Chinese customs data has never shown anything like the numbers that PDVSA reports. I ask industry sources about it, and learn that all it means for “China” to receive the oil is that it’s given to CNPC. They don’t necessarily send it to China. Much goes to Singapore, but some may also go to the U.S., Chile, you name it. We don’t know they are double-counting barrels, but it would be easy.

  8. ow

    BTW, I think the oil production story has been gone over to the point of extreme repetition. It is actually quite boring at this point.

    But if you are in Venezuela and have the capacity and desire for some original and important reporting here is a suggestion: why don’t you look into what is going on with all the other state run industries, and in particular the new ones the government claims to be building.

    The auto and tractor industries set up with Iran, the light industries set up with China, and the big new auto parts factory they are supposedly constructing.

    What exactly is going on inside those projects? Are the producing? Is their production increasing? Are they creating wealth or are they just a black hole for government subsidies? Are they bringing new technology to the country? Are they just assembling kits or our new suppliers opening up? What is the status of the new steel and auto parts factories? When will they enter service? Who will the auto parts factory sell to? Are the now nationalized cement and steel industries doing ok? Better? Worse?

    I believe the fate of Venezuela and the Chavez government lies much more in the answers to those questions than in anything having to do with the oil industry. And for a good reporter willing to do some serious work this is truly virgin territory.

    1. sapitosetty Post author

      A real discussion of Venezuela’s recent document releases is virgin territory. That’s why I’m writing about it. If you’re bored, don’t read it. You’re free to be assignment editor on your own blog.

  9. ow


    I think your point relative to the auditors is a pretty weak one.

    Sure, out of the many thousands of audit reports done each year you can find a handful over the course of a couple of decades where the auditors didn’t note obvious problems (thought not all your examples fall into that bucket). That makes the frequency pretty minimal – after all, I don’t see people going around disbelieving Apples financial statements, or Verizon’s, or Embraer’s or anyone else’s just because some auditor somewhere screwed it up once upon a time.

    I find the assertion that the auditor would be indifferent or even aiding PDVSA misstating numbers to be particularly week. KPMGs name goes on every audit and is risking a multi billion dollar lawsuit should something go wrong. So while a small firm in Caracas may be willing to cover for PDVSA because they get much of their business from them KPMG would not. In fact, given the very public allegations against PDVSA’s numbers I am sure KPMG is making sure every “i” is dotted and “t” is crossed before its name goes on the audits.

    It seems you may have gotten some of your information regarding the reports from other sources rather than reading them yourself. The audited financial statements are in dollars, not bolivares, as that is the currency oil is transacted in. Given that, how many bolivares they get per dollar has no relevance to the veracity of the numbers and there is no opportunity for them to manipulate numbers via the exchange rate.

    Lastly, note that none of the numbers in your post from the various agencies are audited or even sourced properly in most cases (it would be interested to hear how Platts, the EIA, the IEA and the others come up with their numbers, but they never say). A couple of years ago PDVSA invited the IEA to come and audit their numbers to help resolve the discrepancy. The IEA declined.

  10. ow

    I don’t mean to pretend to give you assignments. I just thought it was a good suggestion.

    I am not in Venezuela and not being there (not to mention having a full time job) makes it pretty unrealistic to attempt something of that magnitude.

    Hopefully someone will do it. I am sure it would takes months of work but it would be both fascinating and enlightening.

Comments are closed.