Shut-ins update: U.S. crude production is declining

In last week’s article we warned that the next few weeks and months could be very ugly for crude prices due to fundamental supply and demand factors. Even so, our jaws were on the floor at seeing Monday’s WTI May contract trade at -$37/barrel. As most of our readers are aware, negative prices for the WTI contract for May delivery were largely due to technical factors in physical markets. Specifically, future traders got stuck long the May contract without ability to take physical delivery and had to unload at fire sale prices. While we likely will not see a repeat of Monday’s bloodbath, prices will likely remain under immense pressure due to significant oversupply. The latest EIA data indicates domestic crude production is declining in response to market forces, potentially averting future negative prices. Declining crude production, coupled with anecdotal evidence, suggests that well shut-ins are gathering pace.

“Flattening the curve” for crude storage

U.S. oil producers may be able to successfully “flatten the curve” for weekly injections into storage. As the chart on the left shows, total U.S. weekly stocks of crude oil, excluding the Strategic Petroleum Reserve, stand at over 519 million barrels, a little under the record amount of 536 million barrels in storage set in March 2017. We continue to expect crude stocks to reach a new record by the end of this month. For context, in its most recent report, the EIA suggested total U.S. working storage capacity of around 653 million barrels.

Note, however, the chart on the right. It shows that the pace of injections may be declining, potentially forestalling storage overflows. In the week ending April 17th, U.S. producers injected 15 million barrels into storage, down 4.2 million barrels from the week prior. This is important: if U.S. producers can sufficiently slow the rate of storage injections, they could prevent a day of reckoning, where storage tanks simply cannot accommodate any more volumes.

When we break down the data further, some important findings emerge. In the chart on the left, we see that Cushing storage inventories are rapidly filling up. With net stocks already at 76% of working storage capacity, Cushing storage will likely physically max-out in the next 2-3 weeks (and we suspect all capacity has already been booked commercially). In the chart on the right, we find that crude will likely flow to PADD 3 (the Gulf Coast) after Cushing reaches tank-top: it has the most available remaining storage, and is also situated close to export markets and the U.S. refining complex. In fact, over 40 million barrels have already entered PADD 3 storage this year. One important feature of crude could limit inter-PADD storage transfers: some storage is suitable only for specific grades of crude. Commingling could become an issue as storage builds out beyond Cushing.

While Cushing is rapidly filling out, let’s examine what’s behind the injection slowdown at the national level. Crude oil storage is largely a function of four variables: domestic production, imports, refinery inputs, and exports (some crude is also lost in shipping and handling, but we’ll ignore that in this analysis). The change in storage is equal to domestic production plus imports, less refinery inputs and exports. Let’s consider each variable, starting with refinery inputs and crude exports.

Demand for U.S. crude hasn’t returned

There’s little evidence that domestic and international demand for U.S. crude is recovering. Refinery inputs (left chart) remain at historically low levels and decreased slightly from the prior week. Remember, U.S. tight oil is ideal for producing light end refined products – gasoline and jet fuel – that have taken major hits during the crisis. With refined products increasingly turning to storage at sea, we may not have found a bottom for refinery inputs. International demand recovery is key for U.S. exports, but exports appear to have stalled. U.S. refineries run on a mix of heavy and light crudes (both imported and domestic), but U.S. exports are overwhelmingly of domestically-produced light barrels. U.S. crude exports are also idiosyncratic, tied to discrete shipments (i.e. one Very Large Crude Carrier, or VLCC, can carry over 2 million barrels); week-to-week comparisons are tricky. Still, if demand for U.S. crude is recovering, there’s little evidence of it: U.S. exports were down last week. Demand for U.S. crude remains low.

Imports are also declining

U.S. net crude imports have declined for years due as domestic production surged. Amid the coronavirus, however, reductions in demand are attributable to falling U.S. demand, not displacement from production of light, tight oil. While the U.S. exported less crude last week, it also reduced crude imports by 0.7 million barrels per day to 4.9 million barrels per day, the lowest level since February 1992.

What does this mean for domestic production?

We recognize the EIA’s difficulties of rapidly collecting crude production data at scale from thousands of far-flung, disparate producers. Our analysis of multiple sources of data suggests domestic crude production declined approximately 0.2 – 0.3 MMBPD in the week ending April 17th. Given anecdotal evidence we’ve heard in the market, we suspect that some marginal operations are shutting in wells. Averting a storage blow-out will require even more production declines, and eventually demand to increase. As U.S. crude storage levels climb, there continues to be a very real risk that we could run out of physical storage space for crude, which would necessitate more hasty (and expensive) production shut-ins.