Oil Inventories Build, All Eyes On Balances Going Forward

In our OMF last week titled, "What Happens To Oil Storage In The Next 2 Months Will Play A Pivotal Role For The Rest Of 2024." In it, we wrote:

However, if we are looking solely at oil market fundamentals, refining margins are roofing, and it will be important for the product draws to offset the builds we see coming for crude. If this occurs, then not only will refining margins likely stay elevated through the first part of the year, but it will likely signal an even tighter crude market down the road.

It all boils down to how the rest of Q1 looks. So long as the crude builds get offset by the product draws, oil bulls will be ok. If not, then balances are much looser than it looks.

This week's EIA oil inventory figures were bearish. Crude build was larger than expected as refinery throughput dropped yet again to ~14.5 million b/d versus our ~15 million b/d estimate. BP's Whiting outage resulted in a large w-o-w drop and if we exclude the outage, US refinery throughput will be around ~15 million b/d today.

On a utilization basis, we are now at a 5-year low.

(Note: 2021 saw a polar vortex materially impact refinery throughput.)

Looking at the 4 storages combined (crude, distillate, gasoline, and jet fuel), the product draws were not able to offset the crude build.

And looking at the implied demand figures, EIA had total demand dropping this week by ~973k b/d with gasoline demand falling ~639k b/d w-o-w. We would attribute the large demand drop w-o-w to noise.

Looking at the implied demand from GasBuddy, their figure of ~8.3 million b/d is higher than what EIA reported at 8.168 million b/d. Historically, EIA has reported a slightly higher demand figure than GasBuddy, so we expect this to revert in the following week.

On the total implied oil demand side, we are just in line with 2021 and 2023. Note that EIA has already stated that the 2022 figures are erroneous.

What to expect?

Next week's implied crude storage figure shows another large crude build.