Description
VIC Reactivation: I am applying to come back to VIC after a long period of submitting no ideas. While we are having difficulty finding great ideas, I would like to submit the following for consideration as a reactivation idea. Energy is not our normal place to fish, but we do believe there are interesting ideas in this part of the market currently. I invite you to look back at my old ideas if you need more information on whether to vote for my reactivation.
Company Description:
Antero Resources is a natural gas and natural gas liquids (NGL) producer in the Appalachia basin (Marcellus and Utica shales). In the most recent quarter, Antero produced 2.2 billion cubic feet equivalent (bcfe) per day of gas and liquids (oil production is negligible). The company controls 636,000 net acres in these two basins. Since 2008, the company has never drilled an unsuccessful well. As of the latest reserve report, the company controls 53 trillion cubic feet equivalent of natural gas (~75% gas, 25% NGL).
Company Strategy:
Given the breathtaking advances in drilling and completion techniques during the past decade, production of natural gas in US shale basins (especially the Marcellus and Utica) is more like a manufacturing operation than traditional energy exploration. As a result, Antero’s strategy is driven by manufacturing efficiency and the ability to drive cash margins.
To be clear, this is generally an industry we avoid for one major reason – most management teams are not focused on creating value through the generation of satisfactory returns on invested capital. They seem to focus on elements of running an energy business that do not drive shareholder returns (production growth, acreage growth, etc.) You can sit through hours of energy company presentations at investment conferences and rarely hear about returns on capital. When it is mentioned, it is often discussed incorrectly – as the return on specific wells WITHOUT consideration of company F&D costs, G&A, interest expense, and other real costs. Every company has a presentation slide that says they earn high returns on capital in their drilling program. When you ask why they earn such low company-wide returns on capital, they look at you as if you are from another planet.
While this is an over-simplification, strategic differences in the Appalachia basin essentially boil down to transportation/processing, balance sheet leverage, and hedging. Each company has a slightly different approach to these strategic decisions. In general, the transportation/processing strategy differences among the larger players depend on how the “midstream” assets are controlled. In many cases, the E&P company owns, through a limited partner interest, a piece of the midstream assets. Those LP interests are frequently publicly traded limited partner units. As with all limited partnerships, there is a general partner (GP). The GP usually has an economic structure much like a hedge fund. In other words, GP profits rise faster than LP profits. The E&P shareholders can either own the GP, part of the GP, or none of the GP. They may own LP units in the midstream business, or they may own no “transportation/processing” assets thus relying on third-party services. The transportation/processing strategy drives many E&P decisions. Ownership of publicly traded midstream assets also impacts the valuation of the E&P company, and the volatility of the E&P’s shares.
In the case of Antero, they own a majority of Antero Midstream. Antero Midstream is a publicly traded partnership (ticker AM). Antero Resources currently owns approximately 99 million units of AM. The general partner of AM is primarily owned by the original AR sponsors and management of Antero Resources (recently IPO’d – this is an important economic interest that is NOT owned by AR). This transportation ownership decision is important because it drives some volatility within various E&P shares, and provides additional strategic opportunities. For example, quite often the E&P company will drive capital projects by the midstream company as they build out their drilling portfolio. This can create transportation/processing advantages as infrastructure moves with production growth. Conversely, it can create excess overhead if infrastructure projects are contracted for and built out before production. AR essentially controls the decisions of AM, and thus has some influence over the pace and type of infrastructure that is built out across their drilling footprint.
Contrast this structure with Range Resources which has chosen to own no material infrastructure assets in the region. Rather, they contract with third parties to provide transportation and processing services. EQT shareholders own a significant piece of EQT Midstream, both through ownership of LP and GP units. Midstream cash flows are typically less volatile than E&P cash flows. Thus, share prices of the E&P companies are far more sensitive to natural gas prices. Range is far more sensitive to commodity prices, while EQT would be less sensitive. AR is somewhere in the middle as owners of midstream LP units, but not midstream GP units.
Specific company decisions on transportation contracts and balance sheet leverage drive drilling decisions to a large degree. Most transportation contracts are “take-or-pay”. Simply put, you lock-in transportation capacity, but you must pay for it whether you use it or not (which guarantees the midstream company the ability to service debt used to build out that capacity). The decision to sign firm transportation contracts also impacts realized product prices. As in most ‘normal’ markets, realized price is determined by supply and demand. For example, some long-term contracts with liquified natural gas (LNG) producers are locked in at prices higher than current NYMEX gas. However, you must pay transportation costs to get it from the Marcellus basin to loading facilities on the Gulf Coast. In that example, you may have higher price realizations for product, but your total production costs may be higher given the transportation requirement.
Furthermore, a company’s balance sheet leverage (or transportation contracts) may influence the company’s hedging program. High leverage or significant firm transportation may dictate that a company hedge gas prices for long periods of time, reducing the risk of falling gas prices but also limiting the upside from any increase in gas pricing. While drilling for gas has become more like a manufacturing business, the strategy beyond the drill bit determines a company’s decisions about transportation, production growth, hedging, leverage, and other strategic elements.
Management:
Given the important strategic decisions required in this industry, management is crucial. In the case of Antero Resources, there are several elements of management history that made us interested in studying the company.
Antero began operations in 2004 with two primary equity sponsors – Warburg Pincus and Yorktown Energy Partners. As of the latest proxy, Warburg Pincus continues to own 18.1% of the company. Peter Kagan, a partner at Warburg, sits on the AR Board (since 2004) along with Howard Keenan from Yorktown Energy Partners.
CEO Paul Rady and CFO Glenn Warren have a long history of creating value in energy companies and both are significant owners of AR stock. Mr. Rady purchased 500,000 additional shares in May at a price of $21.11 bringing his personal holdings to ~16.4 million shares (~5% of the company). Mr. Warren purchased 250,000 AR shares in May for $20.99 each bringing his ownership in the company to ~10.5 million shares (~3% of the company). Paul was 63 years old and Glenn was 61 years old as of the latest proxy statement. In my view, Rady and Warren have formulated a sensible strategy for creating long-term shareholder value despite a very difficult natural gas price environment.
Opportunity:
I have followed Antero Resources for several years. However, until just recently we never owned shares because I wanted to see evidence that they thought about capital allocation in a way that seems reasonable. In September, they executed a financial transaction that got little attention, but showed me that they are interested in returns on capital.
Since Antero has a very large firm transportation portfolio, and balance sheet leverage, they have always operated with a very large hedge book. Indeed, the value of their hedge book has frequently been a significant percentage of their equity market cap over the past couple of years. They are also large owners of Antero Midstream LP units. In September, they executed a transaction to monetize both a part of their hedge book and some of their AM LP units. I have argued that they should monetize part of their hedge book for some time, but they seemed determined to keep it in place.
In September, they realized approximately $750 million of cash proceeds from a portion of their hedge book, and they sold 10 million AM LP units for nearly $300 million. The importance of the move is magnified by the fact that they won’t pay any taxes on the huge capital gains generated from these transactions. At year-end 2016, the company was sitting on $1.5 billion of net operating losses. Their basis in AM is nearly zero, so they can monetize some of the NOLs on their balance sheet with this transaction. The company immediately used the proceeds to pay down $1 billion in debt. With this debt paydown, the next debt maturity is in 2021.
While this specific transaction does not materially change the intrinsic value of the company, it does demonstrate that management is working sensibly to improve returns on capital. I have often believed that the first CEO in the energy space that truly focuses on returns on capital rather than empire building will deliver incredible relative results. Perhaps Rady and Warren are spending time considering this option. As a long-term investor that does not wish to make near-term natural gas price forecasts, the idea of management utilizing options to increase shareholder value is extremely appealing.
Model:
I am happy to go further into depth about my AR intrinsic value model in the comments. Below, I have given some of the highlights that help me arrive at my IV. Also, please note that the debt figure in the table above shows $3.3 billion. That is pro-forma for the September transaction I discussed in the write-up and does not include the consolidated debt of Antero Midstream (since AR owns more than 50%). In my IV, I am attempting to value the E&P and then adding the tax-adjusted AM equity value to the E&P business.
My analytical technique in building the Antero financial model was to think about the business as if they produced their existing reserves and then shut down the company. While this is unlikely to be the actual outcome, I think it is an interesting way to think about the market’s expectations regarding the cash flows of the existing asset base. It is up to management to deploy the resulting cash flows on behalf of owners, or sell those cash flows to another energy company for a fair price. As I stated in the previous section of this write-up, it does appear that returns on capital are on the radar of AR management, which makes this quite interesting.
My model anticipates the production of nearly 53 trillion cubic feet equivalent of natural gas over the next 24+ years (through 2041). The first three years are generally in line with management guidance when it comes to production growth and necessary capex to fund such growth. From there, I used development costs and capex spending to drive production increases through 2030. By 2033, production volumes begin to fall along with capex requirements.
Besides production and resulting capex, the primary driver of the model is cash margins per mcfe. This is the key value driver of energy companies in my opinion, so these assumptions are critical. AR is not the reported low-cost producer, but they do achieve very solid cash margins. The reason is that their firm transportation portfolio allows them to realize higher product pricing. Importantly, with production/reserves of 25% liquids, AR is levered to NGL pricing on an unhedged basis which is typically more correlated with crude prices.
Interestingly, the transportation portfolio is a current drag on cash margins. AR has a transportation portfolio that is too large for current production. Given the take-or-pay model, they are spending an extra ten cents per mcfe on transportation they are not using. This is a tailwind to PER mcfe unit costs in the future as they grow production. I have modeled $1.24/mcfe cash margins for 2018. My expectation is for $1.42/mcfe weighted average cash margins for total reserve production.
The primary driver of higher cash margins is lower G&A and interest expense per mcfe. Those costs should remain relatively flat on a gross basis, but should decline significantly per mcfe produced. In 2018, G&A plus interest costs will be approximately $0.57/mcfe. Those costs will fall to approximately $0.20/mcfe at peak production.
I have modeled cost of capital at 8.6%. This includes 11.5% cost of equity (which is ~2/3 of total capitalization) and 3.4% after-tax cost of debt (bond coupons range from 5.0%-5.625% and trade near par). There is no need for a terminal multiple given that this model contemplates the company running out of reserves by 2041. Any cash flows used to increase the reserve base need to beat the cost of capital hurdle to add value to the company. Future land/resource acquisitions will be analyzed on that basis.
AR’s interest in AM is modeled simply as an offset to the company’s current enterprise value. Currently, the AM units are worth ~$2.8 billion pretax. To calculate the EV offset, I have applied a 15% income tax to the AM valuation. The rate is low because AR has such a large NOL with which to shelter gains, and that NOL has not been used in the E&P cash flow math. For all E&P cash flows, I attach a 35% tax rate.
Using the above assumptions, I calculate today’s intrinsic value at $32/share. Using the cost of capital from above, intrinsic value would be $49 by the end of 2022. At a purchase price of $19, that works out to an expected return of 20% annually.
As a valuation check, the above mentioned intrinsic value represents 76% of AR’s current PV-10 value.
Pre-mortem:
In my view, the place where I am most likely wrong is the long-term cash margins for AR as they produce their reserves. This one number incorporates many of the assumptions used in the model (cost inflation, natural gas prices, finding & development costs, etc.) so it is the key figure. The question is whether other operators in this industry will act rationally, or will they destroy capital by destroying cash margins.
The second potential source of error is capital allocation. There are several buffers here that I believe are very favorable. First, management ownership of a significant amount of common stock is very important. This represents the largest portion of their net worth by far, so they are likely to behave like owners. Second, the original equity sponsors are still very involved in this business. Warburg Pincus is the largest owner and a Warburg partner sits on the Board.
As I review investment opportunities, I attempt to identify the type of risks that are present (business, valuation, leverage, or a combination). In the case of AR, I think it is most likely valuation risk. They have recently de-levered the business and have a significant hedge book at much higher prices. However, most of the future cash flows result from margins that are higher than current cash margins (G&A and interest per mcfe are the drivers of margin expansion). If I am off by a material amount on cash margins, then I am significantly off on valuation. Business risk seems low since they have not had a drilling miss since 2008.
We also benefit from a management team and equity sponsors that likely need a liquidity event in the next few years. Since management owns a significant amount of stock (>9%), and Warburg Pincus at 18% of the company’s stock, it seems realistic that they may wish to realize value during our holding period (5 years or so). However, if we are wrong regarding the motivation or ability of management to maximize their returns on capital, we could be wrong here.
Catalyst
No specific catalysts noted.