The rapid fall in the oil price has led to massive turbulence on the oil market and a massive decline in the rig count in the United States. One would expect a decline in oil production resulting from this, especially taking into account the quick decline rates of shale oil. However, the decline in the rig count began in October, and up to now no substantial lower oil production was recorded. The issue is especially troublesome, as data are ambiguous. While the EIA predicts significant oil production falls in June and July (more than 90,000 bpd in each month), other data of the agency indicate that actual production increased by 24,000 bpd last week, adding additional confusion to a complex topic. In this article I use import and storage data to shed light on the issue.
A simple connection
Basically, the idea of storage and its development over time is very simple. The stored capacity at a certain time equals the stored capacity at a time in the past plus stockings minus removals in the period between the two dates. In case of US oil storage the equation looks like this: new storage = old storage + production - consumption + imports - exports. Despite the recent surge of shale production the US is still an oil importer, so exports are relatively insignificant. Fortunately, the EIA provides accurate and precise data for imports, exports, refinery consumption and storage, so it is possible to calculate the remaining variable: production. The EIA does also provide production data, as mentioned above. So, theoretically, the calculated production should equal the reported production or should be at least very close.
Does it make sense historically?
EIA provides detailed data for the last decades. However, for comparison I have only used the last 4 years, from 2011 to 2014. For each year I calculated production based on the equation described above for the first five months of the year and compared it to EIA's reported production. For all data I used the weekly numbers, not the 4-week-average-data that are also provided by the EIA. The following charts show the results.
As one can see, calculated and reported production do not match exactly. For 2011, 2013 and 2014, calculated production is very volatile with significant peaks and troughs. Reported production, on the other hand is more smoothly. For 2012, calculated and reported production have a very similar shape, although the two curves are shifted by roughly 200 kbpd, a deviation of approximately 3%. Similar relative deviations can be observed for the other years, where calculated production is always higher than reported production. Overall, I think it is fair to say that calculated production is a reasonable approximation for reported production. This can also be seen when considering the trend lines. For the years between 2011 and 2014, the slope is qualitatively the same.
Production data for 2015
Now, that I showed the similarity between calculated and reported production, I can apply the same principle for the first five months of this year, as shown in the chart below.
This time, the results are different. While at the beginning of the year, calculated production is higher than reported production, at the beginning of May the trend lines cross. At the beginning of April, calculated production starts to decline. This is very much in accordance with the numbers from the last Drilling Productivity Report (a nice graphical representation can be found here). The rise in US oil production during the last years can be attributed completely to shale oil production and as soon as the important shale plays (i.e. Bakken, Eagle Ford, the Permian Basin) decline, so will total US production.
The results in this article support the thesis that US production is already declining. There are other points that support this assumption. The Bakken oil discount (relative to WTI) has narrowed down to less than $1 from more than $7 just 4 months ago. The declining oil production makes also sense when having a more detailed look at the reported raw production figures. According to the EIA, production in the week ended by May, 15 was 9,262 kbpd. But just in the next week (ending May, 22), production should have jumped by 304 kbpd to 9,566 kbpd? In my opinion not very realistic.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.