August 28, 2017 - 11:30pm EST by
2017 2018
Price: 16.38 EPS 1.69 1.74
Shares Out. (in M): 64 P/E 9.7 9.4
Market Cap (in $M): 1,042 P/FCF 9.5 9.4
Net Debt (in $M): 156 EBIT 140 158
TEV ($): 1,198 TEV/EBIT 8.6 7.6

Sign up for free guest access to view investment idea with a 45 days delay.


Introduction:  I own more units of Cone Midstream Partners (NYSE: CNNX) today at $16.38 than I did when I originally wrote it up in Dec-2015 at $10.  I believe the intrinsic value of CNNX has risen since then.  Fair value based on comps is $21 right now (vs. $15 ½ previously), with upside to $30 (vs. $24-26 previously) and downside to just $15.  I am getting paid 7% while I wait and reinvesting all of the distributions.  

I also believe Cone’s long-term growth potential is more likely to be realized.  The first of its co-sponsors, CONSOL Energy (NYSE: CNX), is much stronger financially than it used to be and has continued to improve the productivity of its wells.  Its other co-sponsor, Noble Energy (NYSE: NBL), sold its acreage to a private equity-backed operator who is much more likely to develop the asset because it is not competing for capital within a portfolio that includes other basins.  

Several of the reasons why CNNX was mispriced in Dec-2015 have been addressed.  The units rose to $25 as the clouds parted, but fell once more as new concerns arose.  Total shareholder return YTD is (25)%.  Those concerns are understandable and addressable, but I believe the selloff has overshot because the energy buyside is in disarray.

There is a noticeable lack of fundamental price discovery within energy right now.  Trading feels like it’s dominated by algos and long/short beta-neutral pods.  The algos seem to drive bizarre behavior in the stocks, whether it’s dramatic reactions to simplistic readings of whether a company “beat” or “missed” on some headline metric without regard to explanatory information in the earnings press releases, or the persistence of correlations that don’t make any sense, long after fundamentals dictate the correlations should have ended.  Long/short pods have so little room for error right now that a single wrong move entails career risk for the PMs.  As a result, there is almost no ability for those investors to take a view beyond the next quarter.  It’s amazing how frequently I have been hearing stories of energy pods that got shut down because of a bad month or a couple of bad trades.  

Traditional long only active managers have their own set of issues that keep them from taking advantage of obvious mispricings within energy.  It’s been many years since you could find companies like CNNX with good balance sheets and double digit organic growth potential trading at 10x free cash flow outside of a recession or an industry downturn.  Long-term investors should be all over this opportunity set, but many are struggling to keep up with broad market benchmarks like the S&P 500 and losing assets to passive strategies as a result.  Over the last century energy’s weighting in the S&P has fluctuated between 5% and 16%.  With a current weighting of just 6%, there are bigger and more obvious threats to relative performance these managers need to worry about, like the FANGs.  Tech currently has a 23% weighting.

CNNX’s trading technicals also create problems for different types of active managers.  The modest float of $350 million makes it hard for long-onlies to build meaningful positions, and the daily trading volume of $2 million is a non-starter for many hedge funds that need to be able to get out quickly.

To illustrate how absurd the price discovery has become with this particular security:  CNNX’s correlation to WTI over the last 12 months is 0.73 and higher than all but one of the constituents in the Alerian even though its earnings are purely a function of
wet and dry natural gas volumes from the Southwest Marcellus and Utica and have nothing to do with crude oil prices.  Its correlation to WTI is higher than the crude midstream bellwether Plains All American (NYSE: PAA) at 0.57.  It’s even higher than Permian royalty vehicle Viper Energy Partners (NASDAQ: VNOM) at 0.45, even though almost 90% of Viper’s revenues come directly from selling crude oil at spot market prices on an unhedged basis.  I would like to hear the practitioners of the efficient market hypothesis explain that one.    


My writeup from Dec-2015 has additional background.

Cone Gathering, LLC (Cone) is a gathering & processing company set up by CNX and NBL in 2011 to gather dry and wet gas volumes produced in the Southwest Marcellus on acreage the two companies held in a JV.   Cone has 455,000 surface acres dedicated to it and the right of first refusal on another 186,000 surface acres.  Only the Marcellus zone is dedicated to it on that acreage, but given Cone’s existing infrastructure on the acreage it’s unlikely a competitor could gather more cheaply from wells targeting the Utica zone.

Cone’s assets are grouped in three main areas on the acreage.  CNNX owns 100% of the main system and 5% of the other two systems.  Overall, CNNX owns approximately 85% of Cone’s run rate EBITDA.  The equity was taken public in Sep-2014 at $22/unit.  CNX and NBL each owned 50% of the G.P. and >30% of the common units.

The upstream JV was dissolved in Oct-2016.  NBL paid CNX some cash and they swapped some acreage, with CNX basically ending up with the land in Southwest Pennsylvania and NBL ending up with the land just across the southern border in West Virginia.   In Jun-2017, NBL sold its acreage and its L.P./G.P. interests in CNNX to companies sponsored by a P.E. firm called Quantum.  The deal is expected to close here in the third quarter.  The Quantum affiliate that now owns NBL’s drilling rights is called HG Energy. CNX is disputing the sale of the G.P. interest, saying they have a right of first refusal option that was bypassed.  CNNX is not a party to that dispute.  (I’ll address the implications of the dispute later.)

Quantum is the real deal.  NBL’s former CEO Chuck Davidson is an advisor to Quantum.  They have sponsored the formation of companies like Jagged Peak (NYSE: JAG).  

Cone gets paid a fixed fee for each unit of production they gather
, and the fees are subject to a 2.5% annual escalator.  Like most of their comps, they do not benefit from minimum volume commitments.  In other words, their earnings are a direct function of actual production volumes of gas, NGLs and condensate.  The specific level of commodity prices is not important.  What matters is simply that prices are high enough to incentivize CNX and HG to grow production on Cone’s acreage.  

Gross EBITDA for the Cone system in Q2’17 was $161 million annualized, of which $23 million is from interests in the system that are still held by the sponsors and have not yet been dropped down to CNNX.  EBITDA net to CNNX’s interest was $138 million annualized.

There is a water gathering and handling business that could also be dropped at some point.  It generates $50 million of EBITDA.  CNNX is not currently in the water business.

CNNX’s cash flows are subject to IDRs.  The highest split is set at the typical threshold of 50% above the minimum quarterly distribution.  The incremental split is currently running at 25%, although the total G.P. take is just 6%, and I estimate CNNX will reach the high splits by Q1’18 at the current 15.0% distribution CAGR.

CNNX is competitive with, and in some cases superior to, its direct peers and the broader Alerian index across a variety of the standard MLP metrics, but it receives a materially worse valuation.  Its balance sheet is strong, distribution coverage is conservatively calculated, it’s low in the IDR splits, and it has compelling growth potential.

Somewhat uniquely, it is covering the distribution and
total capex within operating cash flow.  In other words, it needs no external capital to fund its growth and its capital returns.  None of its direct comps can say that, and only three of the 24 Alerian constituents can say that.

I want to emphasize that Cone’s maintenance capex is accounted for at much higher levels than its Marcellus peers even though there are no material differences in the nature of the assets.  When you adjust for that, distribution coverage would be even higher than 1.5x and the equity would be even cheaper than 9.5x 2017E distributable cash flow.




Let’s review some of the reasons why it was mispriced at the time of my original writeup, and what has happened since then.

  • Falling natural gas prices:  Benchmark gas prices dipped below $2/mmbtu in Q4’15 as activity levels got ahead of demand.  The concern was that gas prices were too low to incentivize continued production growth, and the growth agenda established at the time of CNNX’s IPO would be delayed indefinitely.   

    In response to low prices, the gas-directed rig count fell to an average of 95 in 2016 after averaging 218 in 2015, allowing supply growth to take a breather.  Gas prices have rebounded to approximately $2.90/mmbtu which I believe corresponds with the marginal cost of supply. Cheniere’s investor day presentation has a slide that shows the U.S. has 30 years of supply that is economic at $3/mmbtu.  That figure reconciles with my bottom up analysis of gas producers’ type curve break evens and the depth of their drilling inventories.  This is a price environment that easily allows production to grow within cash flow on CNNX’s acreage, but any price at the industry’s marginal cost of supply should be high enough to incentivize production growth on the acreage because the rock is so good.  

  • CNX’s coal exposure:  CNX Coal Resources (NYSE: CNXC) is an MLP that owns 25% of  CNX’s legacy coal assets.  CNX still owns the other 75%, as well as 54% of CNXC’s common units.  CNXC’s net EBITDA fell from $118 million in 2015 to $77 million in 2016 as the coal consumption and prices went into free fall.  The implication was that the coal business would generate less cash flow for CNX to reinvest in E&P development, and therefore deprive CNNX of volume growth.

    Since then, coal inventories have begun to clear and prices have started to rebound.  CNXC’s EBITDA is expected to rise to $106 million in 2017.  The market environment has improved to the point that CNX has begun the process of spinning off its 75% stake in order to complete its transformation to a pure play E&P company.

  • CNX's balance sheet: At the bottom of the cycle CNX had $3.7 billion of debt and was leveraged 5.5x. FFO interest coverage fell to 3.0x and FFO/debt was barely 11% - and that's with the benefit of cash proceeds from production hedges! Their publicly traded bonds cratered and the yields spiked into the high teens. The concern was CNX wouldn't have enough operating cash flow to fund production growth because so much of the cash flow was getting eaten up by interest expense. 

    Thanks to rising commodity prices and a deleveraging plan put in place by management, CNX was free cash flow positive (operating cash flow minus capex) for five quarters in a row, and they sold a bunch of assets.  Total debt has been reduced to $2.6 billion and net debt/EBITDA is just over 3x with a target of getting to 2.5x.  Consensus estimates are for positive FCF again in 2018.  The bond yields are back down to 5.9 – 6.3%.  

  • Sponsors weren't running any rigs: CNX and NBL laid down all the rigs on CNNX's acreage and for a while they were only tying wells to sales out of their drilled but uncompleted (DUC) inventory. That was merely a stopgap solution to keep volumes growing. Unless you start drilling new wells, production volumes will at some point begin to fall.

    With the rebound in gas prices, CNX is once again running a rig on the CNNX acreage, although NBL never stood up any rigs because they chose to prioritize development of their unconventional assets in the Delaware and DJ Basin. NBL had formed a separate midstream affiliate to support those efforts, Noble Midstream Partners (NYSE: NBLX), which created an additional incentive to focus outside the Marcellus. They took it public in Sep-2016.



  • Gathered volumes have been declining:  CNNX’s volumes were (5.5)% q/q and (9.5)% y/y in Q2’17.  This is not indicative of a problem with the rock or the drilling inventory depth on CNNX’s acreage.  I have run the economics on >100 gas-focused type curves across the industry, and CNX has decades of drilling inventory with recycle ratios in the top quartile of that sample set.  It’s really good.  

    At my Base Case price deck of $3/mcf Henry Hub, $50 WTI, and NGLs at 35% of WTI, the top quartile wells in my sample set generate a well-level recycle ratio of 5.2x.  Second quartile wells are at 3.2x.  Third quartile wells are 2.5x and bottom quartile wells are 1.8x.

    CNX has 3,400 undrilled locations with highly competitive economics:

    • 101 locations with a 5.2x well-level recycle ratio

    • 533 locations with a 4.9x well-level recycle ratio

    • 1,465 locations with a 4.2x well-level recycle ratio

    • 1,177 locations with a 3.6x well-level recycle ratio

    • 123 locations with a 3.3x well-level recycle ratio.

The economics may not be quite as good as what CNX promotes, but either way they are outstanding, and their well economics have shown a nice positive rate of change over the last few years.  

The 3,400 locations represent 56 years of top quartile inventory at their 2017 turn in line pace of 60 horizontal wells, which is enough TILs to grow production 9% y/y.  That compares with 18 years at Antero Resources (NYSE: AR), 33 years at Cabot Oil & Gas (NYSE: COG), 8 years at Gulfport Energy (NASDAQ: GPOR), 16 years at Range Resources (NYSE: RRC), 13 years at Rice Energy (NYSE: RICE) and 6 years at Southwestern Energy (NYSE: SWN).  

The well-level economics on the CNNX acreage that NBL owned looks terrible compared to my sample set.  NBL claims 1,890 locations that by my math have a weighted average recycle ratio of 1.5x.  Then again NBL has not run any rigs in the Marcellus since Q3’15 and has not benefitted from significant improvements in lateral landing placement, drill times, or completion designs over the last seven quarters.  It’s worth pointing out that this acreage, mostly in West Virginia, is interspersed with AR’s acreage, and AR is generating well-level recycle ratios in the 4x range so clearly the rock is fine.

The lack of activity by NBL is the main reason why CNNX’s volumes are falling.  NBL’s Marcellus segment volumes were (8.3)% q/q and (27.5)% y/y in Q2’17.  That will surely change now that HG is operating the acreage.  It would be unusual for a company to buy acreage and not accelerate activity because bringing forward production to increase the present value is a path to value creation that you can control.  In fact, HG has hired a number of former NBL employees that worked on the Marcellus asset, and
a number of their employees are land men or completions engineers according to LinkedIn – exactly the type of people you would employ if you intended to pursue a growth strategy as opposed to harvesting PDPs.  

I believe Cone can handle materially higher volumes on the West Virginia acreage with minimal incremental capex because those two systems were only running at utilization rates of 6% and 37% in Q2’17.

CNX’s Marcellus segment gas production actually grew 7.2% y/y in Q2’17.  It was down sequentially due to their activity cadence but will ramp again starting here in Q3.   The Marcellus accounts for 62% of total production, the Utica is 15%, coalbed methane (CBM) is 18%, and “other” is 5%.  Cone doesn’t benefit from CBM or other volumes.  Nor are they growth assets for CNX.  

Looking forward, consensus estimates for CNX’s consolidated production are +7.2% q/q in Q3’17 and +12.6% q/q in Q4’17. CNX has been guiding to production growth in 2017 and 2018 for several quarters now, and they have been raising that guidance.

    • CNX 2017 production guidance (+9.1% y/y):    

      • At Q4’16 they guided to 2017 production of 415 Bcfe, +5.3% y/y.  

      • At Q1’17 they raised 2017 guidance to 430, +9.1% y/y, and left it unchanged when they released Q2’17 earnings.

    • CNX 2018 production guidance (+24.4% y/y):  

      • At Q4’16 they guided to 2018 production of 485 Bcfe, +16.9% y/y vs. 2017 guidance of 415 Bcfe.  

      • At Q1’17 they raised it to 505 Bcfe, +17.4% y/y vs. updated 2017 guidance of 430 Bcfe.  

      • At Q1’17 they raised it to 535 Bcfe, +24.4% y/y vs. unchanged 2017 guidance of 430 Bcfe.

Goldman estimates CNX will grow consolidated production at a 14% CAGR from 2016 to 2020.  

Cone won’t get 100% of that volume growth for reasons I’ll explain below, and CNX is just one of the operators on Cone’s acreage, but these are the estimates that are available and they are positive.  In spite of this, the market seems fixated on the lack of growth from the NBL/HG acreage as if that is the more relevant indicator and interpreting it as a permanent condition.

It takes 1.5 rigs to maintain production on the acreage CNX operates and 2.0 – 2.5 rigs in total to maintain an organic growth rate equal to CNNX’s 15% distribution growth trajectory.  CNX is running one rig and it seems entirely reasonable to think HG will operate at least one rig.

Bottom line:  The market is treating CNNX like a broken growth story even though its growth potential is as strong as ever and there are clear catalysts to a return to growth with the start of development by HG.  

  • CNX ramped activity outside CNNX’s gathering footprint:   CNX released a lot of nice data about the economics and depth of their drilling inventory at their analyst day last December.  One thing that stood out was their most profitable drilling inventory (the 101 locations that generate a well-level recycle ratio of 5.2x at $3/mcf Henry Hub) is comprised of dry Utica wells in Monroe County, Ohio (just across the southwestern border of Pennsylvania), acreage which is not dedicated to CNNX.  It’s dedicated to RICE’s midstream affiliate Rice Midstream Partners (NYSE: RMP) because it was cheaper for RMP to service by virtue of already having infrastructure there.  Doing the gathering would have been a greenfield project for CNNX and therefore not as cost competitive on a heads-up basis.  

    CNX started directing more activity to this area earlier this year.  The bearish read through to CNNX is it hurts them because those are volumes they won’t get to gather.    

    The reality is that while the dry Utica inventory in Monroe County is somewhat better than the next tranche of locations, it’s not massively better, plus CNX doesn’t have very much of it.  It’s just 101 locations, or the equivalent of 1.7 years’ worth of drilling at the turn in line pace of 60 Hz wells for 2017.  Moreover, 31 of the 60 wells will be Marcellus that CNNX gathers, and some of the 26 Utica wells they’re turning in line are on CNNX’s dedication (i.e. outside of Monroe County).  This shift in activity is not big enough to prevent CNNX from growing its volumes, and this inventory will run out soon anyway.  Monroe County is just not big enough for CNX to build its business on for the long-term.

  • CNNX has failed to win third party business: 100% of Cone's volumes are from its sponsors. Management used to speak in somewhat optimistic terms of its ambitions to win third party business. They are now saying those efforts have been unsuccessful. The market seems to have assumed it's because they're not capable of it for some reason. I have benchmarked CNNX's cost structure against its Marcellus gathering and processing peers and discussed the figures with IR, and it's my conclusion their cost of service is competitive. Apparently third-party E&P companies have either chosen to go with their own midstream affiliates, and so those decisions were not necessarily based straight up on cost of service, or midstream companies like CNNX were being used as stalking horse bidders for E&Ps to evaluate the internal pricing of their midstream affiliates.

    Winning third party business would have been like found money. I never banked on it and CNNX certainly doesn't need it to deliver a robust multi-year growth agenda. Nevertheless, there are still some third-party opportunities the company is working on.


  • Uncertainty caused by NBL’s Marcellus sale:  AR’s midstream company Antero Midstream (NYSE: AM) and RMP are two peers that have given multi-year distribution growth guidance.  CNNX used to give multi-year guidance but they had to pull it back because NBL decided to sell their acreage, and they have recently been giving distribution guidance only through 2018.  It’s understandable why the market would favor companies that provide long-term visibility over ones that do not, but the lack of long-term guidance from CNNX does not equate to a lower quality business or an inability to grow at competitive rates.

    The sale is expected to close here in the third quarter and I would expect CNNX to give updated guidance based on HG’s development plans on the Q3 earnings call.

    The sale is also causing some uncertainty about who the officers of Cone will be.  The co-sponsors get to appoint management.  The CEO, John Lewis, originally came from NBL and the CFO, David Khani, came from CNX.  Their status is unclear.  Lewis may be stepping down regardless of whether CNX or Quantum ends up owning NBL’s former stake in Cone’s G.P.  If CNX is able to successfully assert their right of first refusal and take 100% control of the G.P. it seems logical that Khani would stay on, but if Quantum ends up with NBL’s 50% stake in the G.P. they may want to appoint somebody else.  It doesn’t make sense they would want to make management changes that were disruptive or destroyed value, but the simple fact that key leadership positions are up in the area is possibly a deal breaker for some investors.  

    In connection with the sale to HG, NBL released 37,000 undeveloped acres from the dedication to Cone.  There may be some confusion about what the release means for CNNX.  Most importantly, this is not an indication that Cone’s asset base can disappear from underneath its feet.  37,000 acres is the maximum amount of land NBL was allowed to release from the dedication, and they are not allowed to release any more.  CNX can only release 15,000 acres.

    The acreage is still subject to a right of first refusal.  HG may try to negotiate for lower gathering fees on the 37,000 acres as a condition of re-dedicating it.  It’s not a zero-sum game for HG, though.  That acreage is in one of the asset systems that CNNX only owns 5% of.  A lower fee schedule will mean incrementally lower dropdown values and lower G.P. cash flow at some future date.  

The market wants more gas and demand is shifting in to high gear.  This is great for CNNX because, again, their earnings grow if volumes grow.  Total U.S. demand was 75.1 bcf/d in 2016, having grown at a 1.8% CAGR over the last four years.  Over the next four years Simmons estimates demand will grow at a 4.0% CAGR to 87.7 bcf/d.  The incremental 12.6 bcf/d is expected to come from:

  • 7.7 bcf/d for LNG exports

  • 2.2 bcf/d for exports to Canada/Mexico

  • 1.8 bcf/d for power demand in the U.S.

  • 0.8 bcf/d for chemical production

  • 0.1 bcf/d for transportation

LNG exports is the biggest component of growth but those volumes are not necessarily locked in.  Export contracts are take or pay, i.e. the counterparty is not actually required to take physical delivery of the gas.  They are only required to pay the contracted liquefaction fees.  Henry Hub therefore needs to be able to compete globally in terms of its marginal delivered cost (ignoring liquefaction fees since they are a sunk cost) in order for the volumes to move.  I think the gas will move.  

  • Marginal cost to Europe:  Henry Hub is approximately $2.90/mmbtu.  Variable costs to ship it to Europe are ~$1.50/mmbtu.  Total marginal cost is therefore $4.40/mmbtu which compares favorably with current U.K. spot prices of $6.13/mmbtu.

  • Marginal cost to Asia:  Variable costs to ship it to Japan are ~$2/mmbtu, so total marginal cost is $4.90/mmbtu, well below Japanese spot prices of $6-7/mmbtu.

China has some very nice gas demand trends worth calling out that should be positive for LNG demand.  Total Chinese gas demand in 2016 was 20.3 bcf/d, +7.7% y/y.  The country is planning to raise gas from 6.4% of the energy mix in 2016 to 10% by 2020 and 15% by 3030 driven in part by coal to gas switching (coal was 64% of the energy mix in 2016) as they try to improve air quality.  In other words, gas will take a growing share of a growing energy pie.  China National Petroleum Corp’s base case estimate is for demand of 42.5 bcf/d by 2030.  Domestic production was just 13.4 bcf/d in 2016, so they will clearly need imports.

In looking at bottom up consensus estimates for the top 35 domestic independent gas producers, it becomes clear that >80% of U.S. production growth through 2020 will be supplied by just seven Northeast producers: AR, COG, CNX, GPOR, EQT Corp (NYSE: EQT), RRC and RICE.  No other producers have the scale or rock quality to meet the need.  Without this group, very little of the incremental 12.6 bcf/d demand can be met.

If you combine CNX’s 949,000 acres of Marcellus and Utica stack potential with the 363,000 net Marcellus acres NBL sold to HG, the 1.3 million equivalent acres is the second biggest footprint in the industry behind EQT/RICE at 1.4 million equivalent acres on a pro forma basis.  It’s almost certain that some amount of the industry’s production growth will have to run through Cone’s system.

Let’s review the incentives for CNNX’s upstream sponsors in light of the recent debacle at RMP, where a misalignment of incentives between the upstream entity and the midstream affiliate led to a disaster for the midstream company’s investors.  

RICE agreed to be acquired by EQT for a 37% premium back in June.  As part of the deal, EQT’s midstream affiliate EQT Midstream Partners (NYSE: EQM) got to take RMP’s
future drop down inventory, and RMP got demolished, falling 27% the day the deal was announced.  RMP will probably be acquired by EQM at some point but EQT hasn’t made a decision about how to handle it.  In the meantime, RMP is somewhat of an orphan security.

The growth potential at CNNX from dropdowns is material and it’s important that it doesn’t get taken away.  Between additional interests in the G&P system and the water business, droppable EBITDA is $73 million, and would represent >50% growth over Q2’17 annualized EBITDA of $138 million net to CNNX.

The EQT/RICE deal made sense from an upstream perspective given the complimentary nature of the acreage positions, but it also made sense for EQT from a midstream perspective because the additional dropdown inventory recharges the growth potential at EQM and its publicly traded general partner, EQGP.  

EQM was struggling with the biggest total G.P. burden of the Marcellus G&P peers.  The total take was 32.4% in Q2’17 compared with 20.4% at AM, 5.7% at CNNX and 5.4% at RMP.  

It’s also understandable why RICE was willing to sacrifice RMP to get the deal done because the incentives were aligned in that direction.  Members of the Rice family owned over $800 million worth of RICE and hardly any RMP, and RICE’s $500 million stake in RMP common units represented barely 5% of RICE’s enterprise value.

I don’t believe CNNX will get left in the lurch any time soon.  For starters, CNX is probably not acquirable for at least the next six months while they finalize the coal spin, plus you have two major shareholders who built their stakes at much higher prices than $14-15 where CNX is currently trading.

Southeastern owns 51 million shares of CNX, or 22%.  They bought the first 25 million shares in Q2’12 in the high 20s/low 30s.  They added the next big chunk in Q4’14 and Q1’15, also in the high 20s/low 30s.  

Greenlight owns 23 million shares of CNX, or ~10%.  It’s harder to tell what their cost basis is because they have ramped the position up, down and back up again, but I’d be surprised if they’re in the black because they first started buying in the 30s.  

Management probably wants to run CNX as a standalone E&P company for a while after the coal spin anyway.  CNX’s transformation from a coal producer to a shale gas producer is probably one of the most remarkable transformations in corporate history.  Think about all the companies in industries that got disrupted; Kodak, Peabody Energy, Blockbuster Video, Yellow Pages, Nokia, travel agents, some of the newspapers, etc.  How many of them made a successful pivot to something new?  Hardly any.  The coal business has been an albatross around CONSOL’s neck and I think management wants to show off their ability to demonstrate value creation as a standalone E&P.  I don’t think they’ve ever gotten the credit from the stock market they believe they deserve.    

Nevertheless, I don’t think there is a logical acquirer for CNX.  AR, GPOR, RRC and SWN are the other major players in the Southwest Marcellus/Ohio Utica besides EQT/RICE.  

AR is the only one in the group with a midstream affiliate that could take CNNX’s lunch money (i.e. its dropdown inventory), but AR is well hedged and has plenty of growth runway for their midstream affiliate.  They recently took the midstream’s G.P. public (NYSE: AMGP).  The expected distribution growth rates for AMGP are dramatic (30% CAGR over 2018-21) and yet the shares have performed horribly (down over 20% since the IPO, albeit in a tough energy tape).   I doubt they’d get credit for increasing the growth rate by getting their hands on CNNX’s dropdown inventory, although from an upstream perspective it could make sense for AR to acquire HG’s acreage down the road given the overlapping footprints.  It’s probably way too early for HG to think about exiting anyway.  It may also mean something that AR chose not to buy the acreage when it was available.

GPOR just did a big deal to buy their way into the Mid-Con so they’ve got their hands full.  RRC is already leveraged 3.5x and needs to deliver value from the MRD acquisition they did last year.  Neither of those companies have midstream affiliates though, so presumably CNNX would remain the beneficiary of the dropdown inventory if either of them acquired CNX.  On the other hand, CNNX’s acreage would have to start competing for capital with the other basins GPOR and RRC are in.  For what it’s worth I think the acreage competes very well with those basins.  

SWN would be a more logical acquirer.  Their inventory depth is very shallow and they need to replenish it. They only have 6 years’ worth of locations at their 2017 drilling pace with recycle ratios in the top quartile of my type curve sample set.  Buying CNX would fix that.  

However, SWN and CNX are both leveraged >2.5x net on 2017 consensus so a deal would have to be 100% stock, which raises the question of whether CNX would even want their currency.  I think the answer is no.  CNX already has decades worth of inventory with top quartile gas economics.  If they take SWN paper they would experience massive inventory dilution but they wouldn’t get a better balance sheet or a stronger company in exchange.  (Like GPOR and RRC, SWN doesn’t have a midstream vehicle that could swipe CNNX’s dropdown inventory.)

It hasn’t been determined yet whether Quantum will get to keep the Cone IDR interests they are trying to buy from NBL, or whether CNX will get to buy them.  I don’t know who I should root for.  There are pros and cons to both possibilities.  Either way CNNX is more than cheap enough to compensate for any eventualities.


CNX’s CEO got turned on to The Outsiders and he is constantly talking about growing NAV per share.  He at least says the right things.  So far, his actions seem to back it up.  To that end, CNNX has a chance to become a very material part of the NAV.  CNXC’s enterprise value of $620 million is based on the 25% interest in CNX’s mines.  The implied value for the entire coal business is therefore $2.5 billion, and will go away after the spin.  Take that out of CNX’s current EV of $5.8 billion and you’re left with $3.3 billion for the upstream business and Cone.  By my math Cone could be worth $1.3 billion to CNX, or 40% of the EV.  (The percentage could end up being even greater if CNX has to retain some of the coal liabilities.)  I get to $1.3 billion as follows:

  • Common units ($630 million):  In my High Case (see Valuation section below), CNNX common units could be worth $29, making the 21.7 million units owned by CNX worth $630 million in total.

  • G.P. interest ($310 million):  The G.P. distributions in my High Case after all the drop downs are complete would be $37 million per year.  CNX’s 50% share would generate cash flow of $18 million per year.  Mature IDRs with total G.P. takes in the >30% range are getting taken out by their limited partners at 12-13x cash flow.  In my High Case these IDRs have a 25% total take.  Given the remaining growth runway in CNX’s drilling inventory, a higher multiple is warranted and I use 17x. EQGP is trading at 16.6x on 2017 consensus and their total take is 32% right now.

  • Dropdowns ($310 million):  I value the remaining gathering drops at $290 million and the water drops at $325 million (see Valuation section for the assumptions) for gross proceeds of $615 million.  CNX’s 50% share would generate $308 million.

    This highlights another key difference between CNX and RICE.  RICE was leveraged just 1.5x on 2017 consensus EBITDA, so their debt was not an impediment to a takeout.  CNX is not exactly bloated with debt at 3.0x on a gross basis, but it’s enough to be a roadblock.  There isn’t a lot of appetite for debt in E&P land these days.  CNX still wants to reduce debt, and dropdown proceeds of $310 million would make a good dent in their net debt load of $2.3 billion.  

You can bet management teams have been put on notice that the type of behavior on display by RICE with RMP is an impediment to accessing the capital markets.  I saw the CFO of Antero a few days after the EQT/RICE announcement.  He said he was shocked RICE’s conflicts committee would let that happen and it would be the first topic of discussion at Antero’s next board meeting.  Ironically, even after RMP got its lunch money taken by EQM, it is still trading with a yield 150 bps below CNNX and a multiple on 2017 distributable cash flow that is 2.8x higher.  Ridiculous.

The annual bonuses for CNX management are based on whether they achieve a targeted level of free cash flow, defined as operating cash flow less capex plus asset sales.  The formula basically amounts to cash flows from operating activities plus cash flows from investing activities.  Since cash flows between the midstream and upstream business are probably a wash on a consolidated basis, I don’t think this hurts or helps CNNX for the most part, although drop downs funded by external debt raised by CNNX would probably be net positive to that calculation.

This is one case where I’d actually like to see management compensated on raw production growth!  CNX used to do that but they stopped in 2016.


  • Low Case ($15/unit):  With just the one rig running, they can grow the distribution at the current 15% CAGR through 2018 by eating into coverage and drawing it down from 1.5x currently to 1.0x.  At the end of 2018 the distribution would be $0.34 per quarter, or $1.35 annualized.  That would imply distributable cash flow of $1.35/unit (down 23% from LTM distributable cash flow of $1.74 as of Q2’17).  I assume gathered volumes fall in line with that scenario, at which point they stabilize with the addition of 0.5 rigs.  Again, just 1.5 rigs are required to hold production volumes flat on CNNX’s acreage.

    I think it’s pretty safe to assume the acreage will eventually be developed by at least 1.5 rigs, whether the incremental 0.5 rigs come from CNX or HG. The rock is easily good enough to justify the investment and the sponsors don’t appear to have other material investment opportunities that compete for capital with the acreage dedicated to CNNX.

    A no-growth MLP should probably trade for 10x distributable cash flow, so 10x on $1.35 is $13 ½ per unit.  You will have received $1.28 worth of distributions during the interim (based on the Q2’17 distribution of $0.2922 and a quarterly sequential growth rate of 3.6%).  That’s $14 ¾ worth of value but this is so punitive I’ll just call it $15.  

    The 10x multiple might be too high, but then again it equates to a 9% yield.  MLPs with second tier assets, bad balance sheets, skimpy maintenance capex and weak distribution coverage trade at that level, and none of those characteristics apply to CNNX.


  • Base Case ($21/unit):  I assume the incremental 0.5 rigs necessary to hold production flat gets added immediately and volumes stabilize at current levels, allowing CNNX to put a floor under distributable cash flow at the Q2’17 annualized level of $1.84.  

    The peer group of gathering MLPs with E&P sponsors (AM/EQM/HESM/NBLX/RMP/WES) is trading at 13.5x consensus 2017 distributable cash flow.  Let’s say CNNX is the worst of the bunch and give it a 2.0x discount. Cash flows will grow at the 2.5% escalation rate even if volumes hold flat, so you’re still getting some growth.  A 2.5% growth rate is not that far off the weighted average distribution growth rate for the Alerian of 4.5% for 2016-20.  Put an 11.5x multiple on the $1.84.  That gets you to $21.  Again, the $1.84 is calculated more conservatively than peers.


  • High Case ($30/unit):  This could take a couple years to play out, but that’s okay.  The upside is big and you get paid 7% in the meantime.  I’ll just list my assumptions below rather than copy and paste the whole model.

    • Henry Hub in the high $2s held flat forever

    • CNX and HG run the 2.5 rigs necessary to drive a 15% volume CAGR, in line with the current distribution growth rate

    • The remaining G&P EBITDA of $37 million gets dropped at an 8x multiple

    • The water business with EBITDA of $50 million gets dropped at a 6.5x multiple

    • Maintenance capex runs at 11% of EBITDA for the G&P business and 5% for the water business – pretty robust levels

    • All drops are funded with 100% debt at a 6% interest rate, resulting in pro forma debt/EBITDA of 3.5x.  This is within the company’s target zone of 3.0 – 3.5x and the revolver covenant of 5.0x.  

    • The distribution is raised until coverage shrinks to 1.1x.  

    • Pro forma distributions per unit are $1.75 annualized (50% higher than the MRQ annualized rate of $1.17)

    • The $1.75 gets capitalized at a <6% yield = $30



  • A replay of the RMP disaster:   That deal showed what can go wrong with midstream affiliates when the sponsor gets acquired and the MLP gets sacrificed.  For the reasons I walked through earlier I don’t think CNNX will get its dropdown inventory taken away, but it’s something I can’t rule out 100%.  

  • Domestic recession:  This is the bigger concern.  CNNX makes its money on gas volumes, and gas demand fell in 6 of the last 7 recessions.  

    We’re nearly 10 years in to the current expansion and the average business cycle is typically 6 years.  There are valid reasons why expansions should last longer these days (e.g. better matching of supply/demand across the economy by virtue of access to real time inventory management data, quicker response times in supply chains, etc.), but this one is still pretty long in the tooth.

    I am also concerned about the political situation.  A famous short seller back in the day (Robert Wilson) once said that the most important factor underpinning the market is its faith in the office of the Presidency.  For the first time since Nixon we will probably see that assertion put to the test.  A Constitutional crisis seems like a growing possibility.  This concern is of course not specific to Cone or natural gas demand, but I think it will have broad-based negative effects for the economy.

  • China meltdown: I don't have anything particularly insightful to say about this risk. It seems like China is emblematic of the global imbalances that have developed since the financial crisis. At some point when they get resolved it probably won't be pretty. It could have some negative effects on Cone with respect to lower gas demand, but it will definitely be bad for the energy sector in general. The plain fact is energy is cyclical, and with China accounting for 80% of oil demand growth since the GFC, if they have a crisis the whole sector is going to trade down.



  • The consensus narrative:  CNNX is a broken growth story and is somehow correlated with negative crude sentiment.  

  • My variant perception:  CNNX has nothing to do with crude oil, and its long-term growth outlook has actually improved over the last two years as CNX has figured out how to make better wells at lower costs, and NBL’s half of the acreage ended up in the hands of an operator who will actually develop it.

  • What it's worth: $15 in the Low Case, $21 in the Base Case, and $30 in the High Case. The risk/reward is heavily skewed to the upside with very little downside. I'm getting paid 7% while I wait.


  • Catalyst: CNNX providing updated guidance after incorporating HG's development plans



I do not hold a position with the issuer such as employment, directorship, or consultancy.
I and/or others I advise hold a material investment in the issuer's securities.


CNNX providing updated guidance after incorporating HG's development plans.

    show   sort by    
      Back to top