$ARCH Update: "Let's Have Some Fun" Has Arrived

Looking more like the Bull Case for Met, and the "Holy **** Am I Reading that Right" for Thermal

ARCH reported its much anticipated 3rd quarter earnings Tuesday. At this earnings release, ARCH has normally wrapped up negotiations with its domestic Steel customers and in the past couple of years has had something like 2/3 of Thermal committed. Imagine my surprise as I was speed-scrolling the release and couldn’t find the table that has been in their release year after year.

This table was instead replaced with a high level discussion on the strength of the Met market, and the following snippet on Thermal.

Interesting. In the week leading up to the call, Coronado (~7M ton per year US Met business) disclosed that they have wrapped up their US domestic negotiations at $170/ton for 2022. ARCH has seemingly decided that $170 is an unacceptable price for them, as they later disclosed on the call that they have only committed 400k tons for ‘22 - but at….. $230 per ton! This is quite the bid/ask spread, and is obviously influenced by the fact that current netbacks @ USEC pricing spits out $300 at the mine. I am not sure where this will all shake out, but we can definitely throw out Scenario 1 and 2¹ from my initial writeup where we ran either $105 or $155 Met through the model. I am now moving the base case to $185 for the year.

So that is the Met segment. All signs are pointing towards outstanding outcomes in 2022, albeit with a bit more risk given the company has committed far less volume than years past. While the sell side is asleep at the wheel on ARCH numbers (consensus came into the print at $600ish million for ‘22 EBITDA - this is embarrassing), I would say that there was nothing game changing for those watching the stock closely.

Thermal, on the other hand…..

First, we learned that management has increased their capacity to 70M tons annually in the PRB, a major step-up versus ‘20 and ‘21 volumes, perfectly timed. Second, we learn that the company is sold out on these ‘22 volumes for $16 per ton. Even allowing for a material step-up in the cost structure (1/3 of costs are ‘sales-related’), this means we are printing substantially higher margins than years past. Third, management clarifies that the 70M tons @ $16 is PRB-only, and thus you have to add West Elk on top. West Elk will produce 4M tons and half are priced on Newcastle, which is currently trading, checks notes, 10x higher than PRB. At this point, I am basically falling off my chair… but then it gets better. Management clarifies that they built 80M tons of business in ‘23, ‘24 and beyond at these high prices (while they mentioned backwardation ‘after 22’, remember that they would have already had a lot of PRB tons sold well under the $16 average, so the incremental had to average well above $16 to make the math work).

Let’s just say that I opened my model, looked at what I had for Thermal, and actually fell off my chair.

I will paste my new base case below, as well as the “Let’s Have Fun” scenario from the initial writeup. Take a look at Thermal. In ‘22, I previously modeled $14.50 ASP and $11 of costs on 60M tons. This rolled up to $210M of EBITDA. I was a bit off. The new math looks like this. I believe I was close to $200M too low for ‘22 EBITDA.

Further, I had previously built a two-step model where I threw 7x FCF on Met and 1x FCF on Thermal - this was essentially my way to ‘exit’ the Thermal business. Now that management has built such a substantial book of business at high prices, I am adding 1x $200M of incremental EBITDA to the price target (Think of this as the locked EBITDA they are going to generate in ‘23 and ‘24 if they are contracting at $16).

“Let’s Have Fun” Case from Writeup

So there we are. The bottom line is that Met is playing out somewhere between the $155-$205 ranges I first modeled out, while Thermal has surpassed my wildest expectations on volumes and pricing. I have made a bunch of updates to other parts of the model in the interim (ht to @knlcbd for some great pointers after the initial writeup)². If $ARCH does everything I now expect as my 50/50 base case, I can get the stock to $305. As always, the disclaimer is that buybacks impact the convexity. Same numbers but stack cash rather than buy stock and I show $206. Please note I previously had modeled $100 for the share buybacks and I am now modeling $125. This actually makes a massive amount of difference - if I were to throw $100 as the buyback price, the ending price would be $100 higher. I know, it’s wild. So is a real business trading for 1x NTM EBITDA.

New Base Case

1. $ARCH: Thoughts on the Met Coal Supernova - by twebs - The Asymmetry Infantry (substack.com)

2. I was too low on beginning share count - didn’t account for the convert properly. I was also backing out non-cash G&A. I have updated both (both are a detriment to share price)