Page 1 of 1

Question re 8/23 crude oil storage report

Posted: Wed Aug 23, 2017 3:38 pm
by cmm3rd
Today EIA reported crude oil inventory for week ending August 18 at 463.2 mm bbls (which is again down from the preceding week, and down from peak of about 535 mm bbls).

Dan, you have posted (and I have read elsewhere) that appropriate storage level is said by many to be 400 mm bbls (Raymond James' 465 mm "new normal" figure notwithstanding). However, you have also argued that because consumption is dynamic (fluctuating and, over the long term, increasing), proper levels should be thought of in terms of days of supply.

In that regard, you said in your most recent podcast that 25 days of supply should be viewed as a "critical" level. You then showed a days-of- supply chart which included a line showing the 2011-2016 average, which fluctuated seasonally between about 24 and 27 days, and which for the latest week (again, the average of 2011-2016) appears to be between 24 and 25 days.

So, what is the appropriate number of barrels per day to use (for consumption) in the denominator (to divide into the total inventory level)? Given that the "average" line fluctuates seasonally, I am guessing that the denominator changes weekly. Is that right? What figure do most analysts (especially those who favor use of the "days of supply" metric) use and how do they arrive at it?

According to EIA, the US consumed 19.63 million bbls pd in 2016.

Fwiw, 463.2 mm bbls divided by 19.63 mm bbls/day = 23.6 days, which would be much lower than the 26+ days noted in the chart used in your last podcast (for last week), so my calculation cannot be right. I suspect the reason is that for the denominator I should be using a number other than the 2016 daily average.

Any help would be appreciated.

Thanks

Re: Question re 8/23 crude oil storage report

Posted: Thu Aug 24, 2017 6:06 pm
by dan_s
"According to EIA, the US consumed 19.63 million bbls pd in 2016." < This is total liquids consumed, which includes NGLs. So, not what should be used to calculate dates of supply of crude oil.

U.S. refiners consume 16,500 to 17,500 barrels per day except turning the first quarter when a lot of them do annual maintenance. Crude oil consumption peaks in Q3 each year.

EIA actually calculates the days of crude oil supply for you.
> Go to: https://www.eia.gov/dnav/pet/pet_sum_sn ... _nus_w.htm
> Scroll down about half the page and you will see days of supply listed for crude oil and several refined products.
> The calculation is > Commercial Stocks of Crude oil / Refiners Crude Oil Inputs 463,165,000 / 17,461,000 = 26.526 days, which is where we are now

The "normal" level is 25-30 days. If we get down to 20 days of supply, I would call that "critically" low with some refiners starting to run out of feed stock. I would argue that in a world where the U.S. still relies heavily on imports from places like North Africa and the Middle East, 30 days of supply should be the minimum.

a If you go back to the EIA website above, you can find a chart that shows days of supply going back to 1982. You will see only few times that crude oil inventories got below 20 days of supply.