Description

CNX Resources is a gas focused E&P company with production in the Marcellus shale, Utica shale and with coal bed methane legacy assets, run by a management team that is obsessed with buying back its own shares. In the next 4 years, they believe they can return 7b USD to shareholders through buybacks and leave a company that is in better shape (2.5x as much EBITDA, same leverage) than it currently is. The market cap is currently 3.0b USD. Obviously, if only part of these buybacks are accomplished, shareholders stand to benefit enormously. At 6x EV / EBITDA with 5b in Net Debt, the Market Cap should be at 7b USD, or 133% above today’s level. Depending on the effectiveness of the buyback program, the actual return could be twice that.

While I believe there is more upside than downside to the price of natural gas, I will not try to pitch a long natural gas thesis in this write-up. CNX typically hedges the NYMEX price and basis risk on 60% of year ahead volumes. Their acreage in the US Appalachia makes them one of the lowest cost producers globally in a market that is becoming more global with each LNG terminal that gets built in the US.

The IDRs of CNX Midstream Partners have just reached their final tier where 50% of incremental distributions accrue to the GP. With the midstream company intending to grow distributions to the LP by 15% per year, distributions to the GP will grow exponentially from here, about 4-fold in 4 years.

The opportunity is created by sell-side valuing the sector based on EV/EBITDA multiples, regardless of reserve life and quality. It also seems like CNX’ daily volatility is purely based on volatility in WTI and quants trading it erroneously as an oil stock, while CNX’ earnings hardly depend on WTI prices.

Buyback Potential

There are 2 buckets making up the 7b USD in potential shareholder return: 4b USD in identified disposals and 3b USD in credit that should become available when CNX grows EBITDA from ~800m USD / year to 2.0b USD / year while keeping Net Debt/EBITDA at a constant 2.5x.

Asset Disposals

In its 2018 Strategic Plan, CNX claims to have 4b USD of potential asset disposals that are possible. This can be broken down in 1.6b USD of dropdowns to their MLP CNXM, 0.2b USD in alternative minimum tax refunds and another 2.2b USD of further disposals, of which 0.4B has just been realized with the sale of the Ohio Utica JV Assets.

As for the asset dropdowns, CNX has 2b USD of gas gathering equipment on its books (gross of depreciation, which is minimal). This excludes the CONVEY Water Business which is probably worth 300m USD. While impossible to verify, CNX expects they will forego 200m USD in 2020E EBITDA by dropping down those assets and put an 8x multiple on this, valuing the assets at 1.6b USD. To compare, CNXM trades at EV/EBITDA of 9.5x and every drop down at a lower multiple would be accretive.

CNX recently dropped down 95% of its Shirley-Penn System for 265m USD. This asset has 2018 expected EBITDA of 22-24m USD with significant growth after 2018. This equates to 11.5x at the midpoint.

Given that previous dropdowns have happened close to or slightly below book value (no goodwill on CNXM’s balance sheet), I believe dropdowns of another 1.6b USD to be possible, simply based on the >2b of midstream assets marked as PP&E on the balance sheet.

Looking at it another way, a company valued at 6x EV/EBITDA can sell its assets for 8x EV/EBITDA, use the cash for buybacks, and still earn part of its foregone earnings through increased IDRs. This is the way an MLP is supposed to work when it does not get abused by its sponsor.

The crux of the remaining 2.2b USD is in the 20 year P1 reserve life (expected to grow as a consequence of continued revisions ) and 50 years of resources. Obviously, 20 years of cash flow are worth more than 10 years of cash flow and I don’t think the market credits CNX for its resource depth. Management seems actively looking at monetizing resources and acreage that are not planned to be brought in production in the near term. Asset disposals should have a minimal effect on profitability.

On June 29th 2018 CNX announced the sale of their Ohio Utica Joint Venture Assets for ~$400M. 2019 EBITDAX was 25-35m USD, or 13.3x EV/EBITDA, mainly because of further development potential.

EBITDA growth

While this part is simple, it is the least impressive part of the buyback story as it depends on credit markets to fund growth and buybacks, while this is far from a certainty in an environment where interest rates are rising. CNX intends to grow production by 20% per year, by spending about 110%-120% of EBITDAX on Capex.

I actually believe there is a good chance CNX will disappoint on this growth, depending on credit market conditions. But CNX won’t be alone in this. AR, RRC and COG have all seen shareholder pressure to prioritize free cash flow generation over growth. If the Marcellus producers as a group were to lower their growth targets that is bullish for natural gas prices, something I’m not modelling in this write-up.

For the exact economics I would refer you to the 2018 CMD presentation by CNX [1]. The expectation is that EBITDA can grow from 800m USD to 2.0b USD in 4 years. If CNX were to keep its Net Debt / EBITDA constant at 2.5x, this would unleash another 3b USD in buyback potential.

To verify this claim of production growth, at the moment CNX is charging dd&a of ~0.90 USD / mcfe. However, CNX future development is more efficient than its past development. Capex per mcfe that gets added to proved reserves was 0.45 USD / mcf in 2016 and 0.56 USD / mcf in 2017. The ~1.0b USD of 2019 Capex should add 1.0 – 2.0 tcfe to developed reserves. 2019E production was 600 bcfe. This is very rough math and it depends on production curves how much exactly gets produced in year 1, but it does show that the Capex program is profitably adding to the proved developed reserves and hence production.

Profitability

I summarized the most important financial data below. Most noteworthy I believe is the 0.50 USD / mcfe that you effectively pay for CNX when you strip out the value of their CNXM holdings. You pay 50 cents per mcf of P1 reserves, without putting any value on the significant resource base that can be converted into reserves, and while 58% of reserves is developed and earning a cash margin (based on Q2 2018 financials, excluding hedges) of 1.40 USD / mcf. As mentioned above, Development costs are somewhere between 0.60 USD / mcf (current) and 0.90 USD / mcf (dd&a charge). 2017 Recycle Ratio was 3.3.

Of course there is time value of money, but with such economics it’s hard to go wrong.

CNX has 475m USD in NOLs. Because the amortization of intangible drilling costs are 100% deductible in year 1, CNX does not expect to pay cash taxes for 4-5 years.

Value of LP+GP of Cone Midstream Partners

CNX owns 34.09% of the LP units of CNX Midstream. At the current trading price of 19 USD / unit, the value of the LP is slightly more than 400m USD.

In January 2018, CNX acquired 50% of CNX Gathering from Noble (or NBL Midstreal) for 305m USD. CNX Gathering owns the GP in CNXM and a 95% interest in various mid-stream assets that are available for drop down.

This effective values CNX Gathering at 610m USD, but this includes the 95% interest in the DevCos. Another way to value it is to look at the Incentive Distribution Rights. CNXM has just reached their final tier where 50% of incremental distributions accrue to the GP. With the midstream company intending to grow distributions to the LP by 15% per year, distributions to the GP will grow exponentially from here, about 4-fold in 4 years according to CNX’ own estimates.

Again, those estimates do not seem overly aggressive. If CNXM were to fully distribute its DCF, distributions would be 0.50 USD / share, versus 0.3479 USD in Q3. This would translate to 28.5m USD in GP distributions and 43m USD in distributions to CNX’ LP interest. Putting an arbitrary 10x multiple on this would value the LP and GP combined at 700m USD.

Management

Management’s long-term incentive program is based on the absolute stock price (50%) and total share return relative to the S&P500 (50%). The short term incentive program shifted from FCF to EBITDA / Share, and management makes no secret that the number of outstanding shares will be a significant driver.

Nicholas J. DeIuliis, as per the Q2 2018 conference call:

“Slide 5, I think, is an important slide and it provides a nice overview for some of our capital allocation decisions. As I mentioned during the quarter, we bought back $300 million of our 8% 2023 notes that results in net interest savings of approximately $14 million a year for the next five years. Our leverage for the quarter, assuming our new 2018 EBITDAX guidance, is 2.4 times. And if you pro forma our leverage for the Utica sale, it drops to 2 times which, of course, is well below our prior stated 2.5 times leverage ratio target."

"The right-hand slide – I'm sorry, the right-hand side of slide 5 shows that we've been repurchasing our shares since October of last year, and we've been doing that consistently for each quarter since. We repurchased 5.3 million shares through July 2017. That brings our total to 17.9 million shares bought back since the inception of the program. In total, we've reduced 8% of our shares outstanding. So, if there are questions on whether you should consider reducing CNX's future share count in your models, I think it would be a mistake to not do that. We said that at our Analyst Day back in March, and I just wanted to reiterate that today."

"If we were to do nothing with the $400 million cash proceeds from the Utica sale, our leverage is going to be at 2 times. We've got the opportunity to use our balance sheet capacity up to the 2.5 times target, like we've discussed. So if that 2.5 times leverage ratio is in fact the leverage target you're assuming, don't forget to change your denominator because we're going to put that capacity to good use.”

CNX is one of the few companies that managed to navigate through the 2016 downturn without issuing any new equity. I believe their rigorous steps in selling and spinning off its legacy coal assets and ‘reinventing’ themselves as a low cost natural gas producer to also have been positive signs.

Natural Gas Market

Gas storage is currently 700 bcf under its 5-year average. A cold winter could push prices up meaningfully. Alternatively, for storage to revert to its own average in 1 year, 700 bcf on an 82 bcf/d US market could absorb a 2.5% supply growth in 2019.

Supply growth will come from 2 sources: Permian associated gas and growth in Appalachia production. The latest EIA short-term energy outlook [2] calls for 2019 dry gas production of 87.7 bcf/d versus 82.7bcf/d in 2018, or an increase of 5.0 bcf/d. I believe this could surprise to the downside. Companies such as COG, RRC and AR are facing shareholder pressure to show more capital discipline and prioritize FCF generation over growth. Many companies are still outspending CFO, including CNX, and are relying on credit markets to fund their growth. Higher interest rates are another reason why Marcellus production might have to be scaled back.

Demand growth will come from

• LNG export facilities. Export capacity will ramp up to 10 bcf/d by 2019 or 2020, from 3 bcf / d now. The majority of this will come in 2019.
• ~1.0 bcf/d of growth can come from decreased imports from Mexico / Canada.
• EIA expects domestic demand growth to be flat in 2019.

With both supply and demand expected to grow >6% per year for a commodity with limited storage, prices will be extremely unstable. Every delay in pipeline commissioning or LNG terminal construction could make prices swing either way.

Long term, the market is becoming more global with every LNG terminal that gets built. Marcellus production is the global low cost production. Not a bad place to be.

Risks

• Management might use cash flows for buybacks, reinvestment of the business and M&A. While I believe the company will make net disposals, all M&A is a risk.
• Management might buy back shares regardless of their price. However, at current share price I believe all buybacks to be accretive to NAV / share.
• All other risks associated with a cyclical commodity business.

Conclusion

• Drilling program should grow production 20% per year, bringing EBITDA from the current 800m USD / year to 2.0b USD / year in 2022.
• 4b USD of potential disposals and 3b USD by borrowing against the increase in EBITDA, will most likely be deployed through buybacks.
• The LP and GP interest in CNXM is worth in excess of 700m USD. IDRs have reached their final tier with 50% of distribution growth accruing to CNX.
• This is on a market cap of about 3.0b USD.
• CNX might underperform their production charges depending on credit market conditions. If all Appalachia producers were to do this, this would be bullish for natural gas.
• Natural Gas storage is historically low going into the winter. 2019 will see most LNG facilities starting operation. This should offset the supply growth from the Permian and Appalachia.

Catalyst

- Another 2.0b of asset disposals
- Another 1.6b of dropdowns to CNXM
- 400m USD of cash proceeds from the Ohio Utica JV being deployed for buybacks
- IDRs beginning to contribute significantly to earnings
- EBITDA growth of 150% in 5 years
- Going into a winter with a historically low natural gas storage
- Going into a year where most LNG terminals will start operating