Description

Long: AMR common equity ($18 price target in 2020, 70% two year IRR)

Price target: $18 in 2020

Current value: $6.25

Alta Mesa (AMR) is an integrated energy company formed as the result of a three-party deal in the STACK play of Oklahoma in 2017. Jim Hackett, the former CEO of Anadarko and current Chairman of AMR, created a special purpose acquisition company in March 2017 called Silver Run Acquisition Corporation II that raised $1bn in cash in the public markets. In August 2017 he announced that Silver Run would purchase both Alta Mesa Holdings, a private E&P operator in the STACK, and a midstream asset called Kingfisher operating on the same acreage. As part of the deal the private equity firm with which Hackett is associated (Riverstone) made a direct $600mm investment in Alta Mesa at the deal price of $10.00 per share.

The background of the midstream asset helps explain why we think the security is interesting. In early 2017 Kingfisher midstream (KFM) was considered an attractive acquisition target given its 300,000 dedicated acres in the second lowest cost oil basin in the US. As a midstream company, KFM’s economics are driven by increased drilling and resource production volumes on the acres dedicated to it. However, 120,000 of those acres were controlled by Alta Mesa Holdings, which, at the time, was a highly leveraged E&P company struggling to fund drilling activity. Most potential purchasers of KFM, therefore, were unwilling to underwrite increased volumes from Alta Mesa, which impaired KFM’s potential valuation. Hackett’s solution was to purchase both assets simultaneously and recapitalize Alta Mesa with a large cash infusion to pay down debt and pre-fund an extensive drilling program, thereby ensuring a significant economic uplift for the midstream asset. While the deal structure was complicated, the company’s investment case today is pretty simple: Alta Mesa has a prefunded drilling program that produces 80% well-head IRRs and 98%+ corporate-level IRRs when including the associated midstream economics captured by KFM on the drilling activity. We estimate that resource production volumes will grow by 375% from Q4’17 (before the closing of the transaction) through 2020 when the program is fully ramped. On the fully ramped production profile in 2020 at strip pricing we estimate AMR can generate ~1275mm of EBITDA or 2.1x current enterprise value. We believe the midstream asset alone is worth more than the total enterprise value at today’s prices, and unlike many seemingly cheap energy securities, AMR has limited net debt at 0.7x 2018 EBITDA and falling as the company grows operating profits.  

With such a description, the obvious question is why the stock sold off from $10 in February to $6.25 today while both the commodity price and the energy sector have been rising. The answer is a litany of factors, but which basically can be summed up as SPAC shareholder selling into a market without many natural buyers given the company’s retraction of its previous guidance (in entirety) only a month after deal closing. The headline issue for Alta Mesa in Q1 was that production volumes in 2018 would be lower than management had assumed at the time of the deal announcement last August due to procedural delays in the deal’s closing. This in turn deferred the timing of the recapitalization and the subsequent ramp in drilling activity. The situation was compounded by worse than usual January weather in Oklahoma and a decision by one of the third-party midstream customers to move rigs to newly acquired acreage for HBP drilling, postponing that third party’s activity on KFM dedicated acreage. Kingfisher also lost a new acreage dedication that it had expected to win. Furthermore, an expected restructuring process on yet another third-party customer was delayed, keeping that customer’s rig count on KFM’s acreage suppressed due to a lack of funding. The company’s SPAC-era guidance was extremely detailed, encompassing multiple years of projections across 82 pages of presentation materials. After the deal closed and their overly optimistic prior guidance ran into a wall of delays and bad luck, they replaced their old presentation with a single page of high level guidance addressing only 2018 – and the numbers were materially lower. The company then released a set Q1 earnings figures rendered nearly unintelligible by acquisition accounting several hours late on the day of announcement (i.e. well into market trading as opposed to premarket), in the process triggering a minor panic selloff in their stock and discouraging all past, present, and potential shareholders. For a newly public company, pretty much everything that could have gone wrong went wrong, right off the bat.

While the narrative above reads like a cautionary tale on the dangers of investing in SPACs, it also explains why opportunity exists in AMR’s stock today. Hackett’s involvement and Riverstone’s $600mm investment (at $10) originally piqued our interest in the name. After a number of conversations with private operators, midstream operators, mineral owners and other public companies in the area, we finally came around to the view that AMR’s acreage is “real” – real in the sense of strong IRRs albeit not achieved in the conventional manner. Other operators typically present strong single well results that are driven by large initial production rates but which come with high drilling and completion costs and typically high acreage leasing costs to have the right to drill in the first place. AMR by contrast is a low cost play, with low initial resource production rates offset by very low drilling and completion costs and low operating costs due to their long operating history in the county (e.g. lower leasehold costs, lower operating costs due to salt-water disposal infrastructure and lower/older royalty structures). The IRRs are on par with Tier 1 acreage in other L48 basins, but that is not at all obvious from looking at 30 day initial production rates.

We view the company as Riverstone’s public drillco, in that the returns from the investment are derived from a de novo drilling program rather than from multiple uplift on existing production or flipping acreage on reservoir delineation. Like other private equity drillco structures, Riverstone specifically picked this acreage for its prospective returns after an evaluation of every major basin in North America using three teams of geologists and nonpublic well information as part of their due diligence. Said differently, the quality of the acreage for new drilling is the reason they are there. One can argue that giving some of the anticipated private-to-public midstream multiple uplift to the sellers was “overpaying” (particularly in light of the subsequent public market reaction), but we find it difficult to argue that Riverstone picked rock  geologically incapable of justifying drilling activity for the midstream in the first place.

As AMR invests the Riverstone cash we believe EBITDA will increase significantly due to underlying well returns and associated midstream economics. Using the current commodity price deck, we believe AMR’s combined EBITDA will increase from ~$450mm in 2018 to ~$1275mm in 2020, split between $400mm in midstream EBITDA and $875mm of E&P EBITDA. Hackett is effectively replicating the value creation playbook he used at Anadarko/Western Gas and plans to IPO the Kingfisher midstream business in 2019. We think KFM can command a peer average 10x EBITDA multiple and the E&P business a 6x EBITDA multiple on current year numbers.  For the purpose of illustration, if one were to hold the above valuation multiples and the commodity price deck constant, then the security would be worth approximately ~$6.40 on our 2018 estimates, ~$13.50 on our 2019 estimates, and ~$18 on our 2020 estimates.

In this sense, the market has correctly valued AMR if 2018 were the only available time horizon. To some degree this is understandable with the SPAC presentation now discredited and the company’s March guidance not offering any help beyond the current year. However, we think this short sightedness creates the opportunity in the stock as the issue has been one of timing and not economics - the production volumes have been deferred, not derailed. Using state level data, offset operator results, and unit economics from the midstream business one can see that EBITDA will increase dramatically in 2019 and 2020 if operator type curves are remotely correct. Importantly, the company has the balance sheet to pursue their plan and controls their own destiny in that regard. In the appendix we detail our working assumptions but the main point is that current trading prices imply either a 2.1x multiple on combined 2020 EBITDA estimates, or getting the $875mm of fully ramped E&P EBITDA for free if one values the midstream business at 10x 2020 EBITDA post its planned IPO in 2H’2019.

Downside protection in AMR is derived from the following: Alta Mesa’s drilling program is prefunded, Alta Mesa’s own activity levels drive 80% of KFM forecasted EBITDA, the company has minimal debt, and the oil price breakeven for their acreage is under $30/bbl. In the event of a return to early 2016 market conditions (commodity price collapse triggered by global recession), we find that the security would be slightly cheap at today’s prices simply due to the increase in midstream profits triggered by Alta Mesa’s drilling activity in the next twelve months. Finally, we would note the fact that a well-regarded private equity firm using private well operating data chose to invest $600mm in AMR just two months ago at prices 60% higher than their current trading values.

Catalyst: execution on the drilling program manifesting in quarterly profits.

• 2Q’18 earnings provide the first full glimpse of the company’s profitability even at an early stage of the drilling program - we estimate ~$92mm in quarterly EBITDA versus an extremely messy $50mm in Q1. We expect $130mm in Q3. We also expect the company to issue its earnings press release on time this quarter, which may not sound like much of a catalyst, but given how low expectations are for this company it may well prove to be one.

• 2H’2019 IPO of Kingfisher and potentially increased 3rd party activity on the 180,000 acres already dedicated to KFM. We are not modeling increased 3rd party activity as a base case going forward, but it does present significant upside if realized.

• Increased detail around water infrastructure economics that allows for EBITDA estimates to move higher for the midstream (SWD and frac water).

• Results from initial wells in Major county increases estimates of drilling inventory

• Q4’19 FCF breakeven

Industry concerns vs AMR positioning:

• Oil differentials: Unlike many Permian producers, AMR has no differential issues as they are currently paying $2.15/bbl to truck their production 60 miles to Cushing (i.e. their operating expense numbers already include the highest cost of transportation). Kingfisher midstream recently entered into an agreement to build a dedicated pipeline from their operations directly into Cushing to lower transportation and costs and increase price realization by mid 2019. AMR’s lower API oil should command a premium to Cushing prices as refining demand is greater for their heavier oil given higher middle distillate yields. Currently AMR’s barrels are mixed with condensate barrels prior to marketing which prevents the company from realizing any premium on their production.

• Basin offtake for associated gas production: AMR has firm transport on moving associated gas out of the STACK and faces no risk of shutting in production. 90% of well economics on AMR’s drilling come from oil and NGL realizations, so gas is seen as a cost center that simply needs to be processed and moved out of the basin. Currently AMR is directing their gas to the Midwest to realize better economics than Permian producers sending gas to Waha.

• Principal / agent risk from management incentives: Management in AMR owns 10% of outstanding shares and are highly incentivized to reach the high end of the deal’s earn out targets of $20 per share.

Appendix: Type curves and other criticisms

We frequently hear that investors hate everything in the STACK because state level data is not great and the companies “always disappoint” on type curve revisions downward – the type curve being a styled depiction of the amount of resource produced by a well on average, which many people use to model forward production and operating profits. We would note that other operators in the STACK have been presenting unbound, single well per section type curves which is the reason why the market has been correctly concerned about negative revisions as production transitions from HBP drilling (1 well per section used to hold a lease) to pad development drilling (8-12 wells per section designed to efficiently drain as much resource as possible from a given section). As the number of wells per section grows, the amount of oil produced per well should clearly fall. While the per well returns decrease, the NPV per drilling section (leasing unit) increases. We think it’s appropriate that companies optimize full cycle (leasing) returns as ultimately full cycle returns drive company-wide returns on invested capital rather than just the half-cycle (wellhead) returns after a section has already been leased. However, lease costs are hard to track compared to well production data, so people naturally use IP30s as a shorthand for half and full cycle returns.

Source: EOG Q2’18 presentation

Unlike other operators, Alta Mesa does not present a modeled type curve based on single well per section HBP drilling but rather on the arithmetic average of all their actual well production data. Alta Mesa’s drilling has historically been done at very tight intervals (as close as 660 ft) which generated low per well results during early historical spacing tests. Their current spacing is 1500 ft (4 wells per bench per section) which has yielded results at the type curve even as they have transitioned to multi well pad development. Well results are outperforming the type curve on a single- well basis because current well spacing is actually wider than the spacing used historically, in addition to the company moving up the learning curve with regard to optimized well completions. On a multi-well basis, the company argues that per well metrics are in line with their type curve (per the below).

Source: AMR June 2018 presentation

Alta Mesa has done the most multi-well development pads of anyone in the normally pressured STACK oil window. We specifically note “normally pressured”, as the well costs and production profiles on their acreage are sufficiently different from the volatile oil and gas windows one county over as to be functionally different resource plays despite all being labeled “STACK”. Hence comparing a type revision by a CLR or DVN in the volatile oil window nearby is not a readthrough to AMR’s results since the rock is completely different. Similarly, the manner in which the parent single wells were drilled also has a significant impact when comparing subsequent child well performance. A huge parent well (large completion) in the middle of a drilling section will present much greater degradation in subsequent child wells because 1) the initial well was large in the first place which makes comparisons worse and 2) a large frac, particularly in the center of a section, will have the most impact on the reservoir for subsequent drilling compared with a small frac along a lease line. Basically it’s nearly impossible to accurately read across from spacing test outcomes in one part of the STACK to predict child well performance in a different county on different lithology in a different subplay of the STACK but everyone does it anyway.

Having said all that, it does seem that a 20% degradation in per well results is “about right” as it stands today when moving from single well HBP drilling to multi-well development drilling, which is offset by 10-15% costs savings per well. Our best independent estimate of AMR’s results is that single wells are coming in 10-15% higher than their published type curves and multi-well pads 5-10% lower on a per well basis. Hence, one would conclude that with 80% of drilling activity occurring on multi-well pads going forward, the type curve for modeling purposes should be 94% of what the company says it is. The issue is that the independent state data from OK which we are using to base our estimates has historically understated actual production in the first six months of a well’s life, and they just started the pad drilling program post recapitalization in Q4’17. In any event, the state data only has about half of the multi-well pads while the company shows the full 14 it has done in the last 8 months. All we are saying is that it’s possible that the company is not misrepresenting their data in the public presentations. However, even if production on multi well pads does come in 10% below the published type curve over time, we would note that AMR has the ability to increase production by using electric submersible pumps versus the gas lift on which their type curve is based. And, at the current valuation an investor is paying $0 for the E&P segment anyway - the returns at 90% of type curve more than justify the activity, which drives the midstream volumes. In order for drilling to be uneconomic, wells would need to produce 40%+ less than the published type curve, which would represent a colossal failure on the part of the multiple geology teams who evaluated this acreage prior to the deal.

Our EBITDA forecasts in the writeup are based on the published Osage type curve with 350 bbl/d peak production and 250 MBO EURs, which is 100% of the company’s published curve and 3.8mm D&C capex per well. We chose the Osage because that curve ties the most closely with the state level data, as opposed to the Meramec which has higher initial production rates. One could argue that an investor should just use 90%-95% of the type curve given pad drilling going forward but then in theory one would also use ~3.5mm D&C capex given 10% cost saves on pad drilling. It doesn’t make a ton of difference.  Ideally, we’d like to know the full NPV of a given section using lease acquisition costs, seismic data, allocated corporate overhead and the full NPVs of all wells in the section to assess the company’s value add. Maybe they’ll present it going forward; we’ve asked them for it. Yet again, an investor today is not paying up to access the returns on the acreage the company leased years ago since the stock trades at tangible book value.

Regarding Marathon’s Q4’17 comments on well level IRRs in the normally pressured oil window, we believe they were primarily in reference to their drilling costs in the Eve test vis a vis resource production. Specifically, the well results on a resource production basis were in line with those produced by AMR but the drilling costs were significantly higher. As Marathon lowered costs in the subsequent Cerny test to $3.8-4.0mm (similar to Alta Mesa) their returns have significantly improved.  Furthermore, Alta Mesa’s highly contiguous acreage position and long presence in the basin has allowed them to invest in SWD infrastructure, which lowers AMR’s net lease operating expenses and increases IRRs compared to Marathon’s offset activity.

https://newsok.com/article/5602102/water-fight-heats-up-in-kingfisher-county

Kingfisher county water issue: the article above triggered a 10% selloff in AMR’s stock during July as it notes that temporary water pipes are facing local regulatory pressure, which will result in “delays and millions of dollars in unexpected expenses” for all E&P operators in Kingfisher county. The article specifically cited NFX but since AMR only operates in Kingfisher county investors assumed the issue had to be the most pernicious for them. The reality is that AMR has built up extensive saltwater (produced water) disposal infrastructure over the last 20 years, which is one of the advantages of having an extremely contiguous acreage position and long lease operating history. AMR uses permanent pipes to transport the saltwater in question, unlike the temporary pipes used by NFX (basically above ground hoses carrying frac chemicals). AMR does use temporary pipes for supply (fresh) water, but that is not in question as the county doesn’t mind clean water being transported above ground in public right of ways. While the regulations in the article have not yet come to pass, if they did we think if anything they would actually help the Kingfisher midstream business. Private equity firms operating in the county do not have the water infrastructure in place to comply and would likely need to hire KFM for saltwater disposal.

Gastar restructuring: on 8/1 Gastar announced that it is actively considering potential restructuring transactions, which likely include some form prepackaged bankruptcy as evidenced by the trading prices of the equity and preferred stock. Gastar is one of KFM’s third party midstream customers, but was only running 1 rig on their acreage. We would view any restructuring as a positive for KFM as new equity owners (via a credit bid or otherwise) would likely try to maximize value by developing GST’s acreage after the restructuring is finished. The worst case scenario is the current one with GST limping along with only one or zero rigs and a severe capital structure overhang.

Catalyst

• 2Q’18 earnings provide the first full glimpse of the company’s profitability even at an early stage of the drilling program - we estimate ~$92mm in quarterly EBITDA versus an extremely messy $50mm in Q1. We expect $130mm in Q3. We also expect the company to issue its earnings press release on time this quarter, which may not sound like much of a catalyst, but given how low expectations are for this company it may well prove to be one.

• 2H’2019 IPO of Kingfisher and potentially increased 3rd party activity on the 180,000 acres already dedicated to KFM. We are not modeling increased 3rd party activity as a base case going forward, but it does present significant upside if realized.

• Increased detail around water infrastructure economics that allows for EBITDA estimates to move higher for the midstream (SWD and frac water).

• Results from initial wells in Major county increases estimates of drilling inventory

• Q4’19 FCF breakeven