I spent some time this morning dissecting the internals of United States propane inventory levels. No, I don’t need more hobbies…
The industry has built 26.0M/bbls of new inventory since April 1st through July 31st, or the first four months of the traditional six-month inventory building period which typically ends in September.
Last year, we built 27.1M/bbls during this same 4 month time frame (April 1st, 2016 through July 31st, 2016).
Here is a look at some recent April-July build totals, beginning with our current year, in millions of barrels:
That equates to a six-year April-July average of 28.6M/bbls.
Even as we take into account June’s record one-month build of 13.5M/bbls, we are still lagging behind last year’s build total and well behind the monstrously large first four-month build totals seen in 2015 and 2014.
Let’s go back a few more years, or pre-shale revolution years. Again, these are April through July build totals for the year listed, in millions of barrels:
This creates a 10-year average of 27.8M/bbls.
Our build during this year’s April-July cycle is below the 10-year average and ranks as the 5th smallest build out of the bunch, or a middle of the pack build.
This year’s build also comes during a time where the export arb has not been as favorable as it was during the winter and last fall, and our exports have been nowhere near the export capacity we now have in our industry.
One trader pointed out yesterday that our present inventory snapshot is ‘eerily similar’ to the same time period of 2013. Any reference to the summer/fall of 2013 will make me take note and do a little research, as that period was the precursor to the most volatile price spike in propane history, which came a few months later in January of 2014.
Here is a side by side look at 2017 inventory levels compared to 2013. I have national inventory level listed first, then the two primary trading and storage hubs broken out below, with the end of July inventory levels highlighted:
The National level shows 2017 being 6.3M/bbls ahead of 2013 at the same time. Production levels now vs 2013 are roughly 500,000/bpd higher, but our export game is much stronger. Propane exports were 252,000/bpd for the EIA weekly report ending 8/2/13, for the sake of an historical baseline. This past week, exports were at 827,000/bpd. My point is a 6.3M/bbl difference in national inventory levels between now and the same time period of 2013 isn’t a big difference at all, a week to 10 days worth of healthy exports and it’s a wash.
Midwest levels are nearly identical; 21.9M/bbls now vs 21.1M/bbls in 2013. The price spikes of January 2014 were most severe in the Midwest, via the Conway trading hub. The Gulf Coast, or Mt Belvieu levels are near enough, with just 3.6M/bbls worth of difference.
So are these levels ‘eerily similar’?
They are close enough to cause some pause, and I have been sounding the sirens of concern over inventory levels for the past several months, pursuant to where I believe we will end up and how we could be setting the table for some price volatility for this coming demand season.
If you take a look at the National section of the chart above and add up the August through September build totals, we added just 5.7M/bbls of inventory over the final two months of the 2013 build cycle. I don’t believe we are in for a repeat of such meager builds for our next two months, but we need to build another 12.4M/bbls to reach the 80M/bbl mark. We have eight more inventory reporting periods left before September ends, and that would be an average weekly build of 1.55M/bbls to reach the 80M/bbl mark.
That is doable, but not a given from a historical perspective.
I remain of the opinion we will come in somewhere around 84M to 86M/bbls by the end of September. To get to that midpoint of 85M/bbls, we would need to average 2.17M/bbls worth of builds per week for the next eight weeks. That might be a high-side stretch, in my opinion.
Through July of 2013, the industry had 45.07 days of supply on hand, by our reckoning. Through July of 2017, we have 40.45 days of supply on hand. By the end of September 2013, we had 47.40 days of supply on hand. We are projecting an end of September inventory level of 84.7M/bbls, which equates to 47.89M/bbls of supply.
Eerily similar? If these things come to pass, and we build another 17.1M/bbls of inventory over the next two months, which is an average weekly build of 2.14M/bbls, yes…we can say eerily similar….or nearly identical.
All the industry would need is a spark to set the volatility flames alight. We had that in 2013 with a big grain dryer year, early onset cold that did not relent and then multiple Polar Vortex outbreaks on the front half of winter. Click here to read my ‘Propane Meltdown’ article from 2014 to relive the steps.
I am not saying we will experience any or many of those things this year, as it is just too soon to tell and the fall of 2013 into the winter of 2013-2014 was a perfect storm of bullishness on top of a precarious inventory position.
What I am saying, again, is the stage is being set for price spike risks, especially in the Midwestern markets tied to Conway hub pricing.
If things get tight in the Midwest and prices move up, shipments of Conway barrels to Mt Belvieu will slow to a standstill, then the Mt Belvieu markets could tighten up if exports remain strong and less product flows south to meet that demand. In that scenario, Mt Belvieu prices would feel upward pressure as well.