EIA Surprise

Jim Brown
 
Printer Friendly Version

There was another "surprise" inventory report today. However, the surprise was EIA numbers roughly inline with what the API reported Tuesday night. The API reported a build in crude levels of 6.4 million barrels and the EIA reported an increase of 5.0 million barrels. Compared to the disagreements common between the two reports this one was at least close.

The API said gasoline inventories fell by -1.8 million barrels and the EIA bumped that to a drop of -4.4 million barrels. Distillates declined -1.6 million barrels.

This is yet another case of different data sets, different cutoffs but the same overall picture over the long term. While the EIA and API can disagree for weeks at a time they always converge again for a brief period of agreement.

Several "reporters," a name that fits them better than analysts, were perplexed by the sudden surge in gasoline demand. I have been telling you for a week that gasoline was being exported to Europe to compensate for the 11 closed refineries. I thought maybe those reporters failed to connect the dots.

After spending an hour researching the inventory history I found the data is not that easy for a reporter to interpret and there were some interesting numbers on the EIA website.

For instance the EIA shows actual U.S. demand rose to 9.358 mbpd for the week. That was up from 8.891 the prior week and 8.812 the week before that. Evidently gasoline demand appears to be rising in the U.S. but you may remember from yesterday the MasterCard Spending Pulse report showed the exact opposite. The report showed that gasoline demand FELL last week by -1.7% and -2.7% over the same week in 2009.

The question then becomes "how does the EIA calculate their demand numbers." Because the EIA numbers comes from the refining sector I suspect the demand calculation is just that a calculation. Inventories at the start of the week were X, plus production of Y, minus the inventory levels at the end of the week.

For example if inventories started the week at 100 mb and refiners produced 20 mb your would have 120 mb before consumption. By subtracting the inventory level at the end of the week, say it was 95 mb then you would assume consumption was 25 mb for the week. There is nowhere in the calculation to say where those barrels went. Did they go to retail dealers, exports to Europe and Latin America, etc. It is a number exercise by a pencil pusher.

There is also the difference in the consumption period. When it leaves the refiner for parts unknown it is consumed as far as the refiner is concerned. When MasterCard totals up the charge slips for the week is true consumption. When you pump it into your car it is no longer in the system. The difference in time from when it leaves the refiner to becoming gas in your tank could be weeks. Some of it travels by pipeline, tanker truck, etc. It then languishes in the tanks of the local wholesaler for some period of time before being transferred to the underground tanks at the retailer.

Obviously it is entirely possible for the refiners to report a huge build in gasoline inventories (at the refiner) while MasterCard is reporting a huge spike in demand. The two don't correlate.

The one we really care about is the real time consumer demand number from MasterCard if we are gauging the economy and rate of consumer demand. The EIA and API numbers are a different time frame and gauge the longer-term demand picture over months instead of weeks.

The image below is from the EIA. It shows the gasoline demand at the refiner over the last year. The blue dots are the 2009 levels. The red line is current demand. The orange line is a smoothed representation of monthly averages. That takes out the individual spikes from things like hurricanes and holidays.

Note that demand at the refinery level is very close to slipping below the 2009 levels. I find it very hard to understand how reporters would say that demand is surging.


In this chart the gasoline inventories were surging back in August but have declined significantly in October. There is a reason for this. The summer blends have to be used up ahead of the switch over to winter blends. The summer blends have different additive ratios to prevent smog in a hot environment. The winter blends reduce the amount of those additives and change the ratios because smog is not as big a problem in the winter. Many states and even local towns and cities have rules about what gasoline must contain based on the month of the year. This forces refiners and retailers to allow stocks to deplete as the seasons change. This happens when the refiners shut down production for seasonal maintenance and the blend changes. This is why the gasoline inventory level declined so significantly. It was not U.S. demand although I am sure some if it is on the way to Europe.


One last chart. This is the EIA chart of crude inventories for the last year. Note that crude inventories are at the high for the YEAR. Last week saw a pickup in the pace of imports by nearly 1 mbpd. They went from 8.6 mbpd to 9.463 mbpd. This is very troubling for the bulls. Yes, there is low refinery utilization right now at 83.7% but that is due to the seasonal maintenance and blend change. Even with the maintenance gasoline production has risen slightly over the last three weeks.

The rise in inventory levels to 366.2 mb and closing in on the 22-yr high on May-1st 2009 at 375.3 mb makes it hard to be bullish about oil. I do believe global demand is rising but these are U.S. inventories and the WTI is priced based on U.S. inventory levels.


I know all of the brokers like JP Morgan are calling for $100 oil in 2011. Some even believe we will see $90 before year-end. As an investor I would like to see that but as an analysts we are going to need to see U.S. demand accelerate before it is going to happen.

Jim Brown

Archives:200920102011201220132014201520162017