I am long shares of Dorchester Minerals (DMLP). I believe there are several pathways for a better than average investment outcome, and comparatively few scenarios that result in underperformance. Perhaps most importantly, I think there is negligible risk of any permanent capital impairment thanks to what I estimate is: 1) an enterprise value well below asset value, 2) zero debt, 3) minimal G&A overhead 4) tremendous alignment of interests with management and 4) substantial cash flow that is being returned to shareholders every quarter. The upside scenarios should come through long-term distribution growth (with a meaningful potential for yield compression) and/or monetization of certain assets.

Taking the distribution growth scenario first, though my estimate of future distributions is anything but exact (I will explain why later), I think it’s reasonable to expect +10% distribution growth for a long time to come. Building this distribution growth on our current yield of 6% gets me to my “base case” expected return of 16% over a multi-year holding period. When Canadian sellside analysts valued one of DMLP's Canadian peers they priced it off of a 7% expected return. I think it is certainly possible that DMLP could see 100-250 basis points of yield compression, to trade more in line with its peers (the capital requirements and growth of these cash flows would support comparisons to an MLP GP yielding <4%), which would result in an additional 20-70% share price appreciation.

Upside also could come through monetization of certain high-growth assets in the Permian, Bakken and, potentially, Delaware Basin. DMLP is trading at a very steep discount to its closest US peer, Viper Energy (VNOM). While I don’t think an acquisition of the entire company is likely, I do think it is possible that VNOM acquires some assets with the result being significant price appreciation for DMLP.

What is Dorchester Minerals?

Dorchester Minerals is an MLP that owns royalty interests (no capital participation) in 378,000 net mineral acres (2,308,000 gross) and working interests (participate in well costs) in 860,000 gross acres (net position not estimated). In 2013 78% of distributions came from royalty interests. The assets are spread over 574 counties in 25 states. Their position in the Permian Basin is large and very high quality. I still can't figure out how to paste graphics, but I would suggest taking a look at slide 11 of Apache's most recent presentation (http://files.shareholder.com/downloads/APA/3412853534x0x760162/bca16559-115b-4047-a1bc-7e30c63704ee/Apache_Bernstein_20140528.pdf). Their Bakken position is also high quality, but much smaller. I would suggest looking at Continental Resources' investor presentation (http://phx.corporate-ir.net/External.File?item=UGFyZW50SUQ9NTUzMTY4fENoaWxkSUQ9MjQ4NTkxfFR5cGU9MQ==&t=1). Their acreage in the Hugoton, Fayetteville and other dry gas plays is not exciting at current gas prices, but if we see +$5 gas then they could become much more valuable.

Unlike most other MLP’s, Dorchester doesn’t acquire much and there are no general partner incentive distribution rights. The last acquisition completed was in 2010 when the company extended its ownership interests in the Permian and Bakken. The deal was done at under 12x run rate cash flows and increased exposure to two premier plays. Though I don’t have specifics on the net acreage acquired, the return on this investment has been outstanding (likely +10x). This was probably more lucky than good, but still impressive.

For the most part, DMLP’s business is negotiating leases with operators who want to drill on their land. After a few conversations with the CEO, I have two take-aways: 1) the team is well educated, both with regard to leasing strategy and E&P development, and 2) the CEO subscribes to what we refer to as “time value arbitrage”, often being the very last mineral owner to enter into a lease and frequently opting to go “non-consent” (meaning they don’t participate in the wells being drilled) until a play is well-defined. While this approach will likely result in the company missing out on bonus payments for hyped plays that fail to deliver and taking longer to show production growth, in the long run we think this is a good approach and is supportive of our view that management is doing an adequate job managing acreage resources and allocating capital when the company elects to participate in drilling wells.  Since coming public in 2003 DMLP has compounded at 17%.

Dorchester’s disclosures are very thin. I don’t think this is likely to change anytime soon. Since the company has no plans to access the capital markets and any additional information they provide can, at worst, hinder lease negotiating processes and, at best, create additional work for their staff of 21, I’m not sure I really blame them. Though limited information creates some analytical challenges, it’s likely one reason for why the opportunity exists and I think there is certainly enough there to get us comfortable with an investment. The following summarizes the most salient publicly available stats about the company’s acreage and production composition.

-        Production

• 2013 Total production of 12.6 Bcfe, down from 12.7 Bcfe in 2012
• 75% natural gas, 25% oil and NGLs, from an 80/20 mix in 2012
• Year over year, oil and NGL production grew 24% and gas production declined 6.9%
• By play:
  • Production is 24% Permian, 23% Hugoton, 16% Fayetteville, 7% Bakken, 6% Barnett, and remaining 24% is a mix of smaller, generally uninteresting plays
  • In 2013 Bakken production grew by 73% and Permian production grew 8%, while production from other basins declined 7.5%

-        Acreage (sorted from most interesting to least)

• Permian Basin
  • Midland Basin – 22,207 net acres and 155,469 gross NPRI/ORRI acres (no idea what the "net" interest is from these)
  • Central Basin – 35,603 net acres
  • Other Permian Basin – 17,917 net acres
• Bakken
  • 8,905 net acres located primarily in Williams, Burke, Mountrail, Divide, McKenzie and Dunn counties
• Other Plays
  • 3,000 Net Acres in New Mexico with Bone Springs, Delaware Basin and San Juan potential
  • 11,464 net acres in Eastern Arkoma Basin

Production and Distribution Growth

Estimating future production and distributions is a challenge for a couple of reasons.  Besides not having much information about the production composition by play or details on ownership percentages in each section, we also have little insight into leasing/drilling activity and the potential impact of previously drilled, but not yet contributing, “non-consent” wells. When Dorchester elects not to participate in a well or well program then the operator receives a return of capital (and in some states a modest return on capital) before the non-consenting mineral owners start to receive their pro rata share of the well proceeds. Nonetheless, one can do some rough math to gain comfort that: 1) production (or at least liquids production) and distributions are growing and 2) this growth should continue for a long time.

By using average realized oil and gas prices in 2013 and typical oil/gas/NGL breakdowns for each play, I estimate that roughly 55% of revenues came from Permian and Bakken production and 45% came from the legacy dry gas production. Though this is admittedly very simplistic, I think it is reasonable to assume that Bakken and Permian production will grow at an average rate of 25% per year for the next several years and the legacy gas production will decline at 6-7% per year. Because there is little in the way of shared well expenses or corporate opex, I think the weighted 10.5% revenue growth should come reasonably close to approximating distribution growth.

Perhaps more important than estimating the magnitude of distribution growth, it’s necessary to determine how long distributions should grow. Again, you have to do some approximating.  In VNOM’s prospectus the company notes that Diamondback estimates there are 229 horizontal locations on the company’s 14k gross acres, or approximately 10 wells per section. Pioneer Resources, one of the largest Permian producers and most significant operator on DMLP’s acres, estimate total recoverable barrels per well between 575,000 and 800,000. This suggests that there could be between 200 million and 275 million barrels of oil equivalent on DMLP’s Midland Basin net acreage. Apply an average royalty of 25% and you get net revenue barrels of 50-70 million. Based on DMLP’s run rate Permian production of 500 boe/d, or 182,500 barrels per year, I feel very comfortable underwriting +10 years of substantial growth.

I use a similar approach to estimate Bakken growth. Drawing from Continental Resources’ +600 MBOE EUR estimate and 2 wells per section, I estimate +4MM barrels of net revenue barrels. While I don’t have quite the margin of error for +10 years of growth in the Bakken, I think it is certainly reasonable to assume several years of robust growth rates.

 Valuation

As I noted in my opening comments, I own shares of DMLP because I think I can comfortably underwrite a +16% expected return for the next several years. That being said, I also own DMLP because I think it is very cheap relative to its peers and I believe it is trading at a discount to its asset value.

While there are several royalty trusts and companies with significant mineral interests, I think the two best comps are the aforementioned Viper Energy (VNOM) and the recent Encana spin out PrairieSky Royalty (PSK) due to the asset quality, production growth runway and capital/expense light nature of the royalty cash flows. Relative to these comps, I estimate that DMLP is priced at a 20-70% discount.

VNOM is priced at an extraordinary valuation because it will grow quickly over the next few years. However, it has a limited runway of production growth, hence its declared interest in acquiring additional mineral interests. Considering the fact that VNOM is valued at +$180k per net mineral acre and their acreage is adjacent to much of DMLP’s Permian position, I think there is a reasonable chance VNOM acquires some, or all, of DMLP’s Midland Basin position. 

While PrairieSky owns a staggering 5.2 million mineral acres, the prospects of the plays on their land are not compelling and it is difficult to estimate much, if any, growth.

In regard to my claim that I think DMLP trades at a discount to its asset value, I put together the following sum of the parts walk.

While I am certainly happy to discuss each of these estimates, I think the point of this exercise is to establish that the assets offer substantial margin of safety.

Alignment of Interests

The CEO, CFO and COO each make $115,000/year in salary. The CEO, Casey McManemin, owns 1.2MM units so his LP distributions are 21x his salary. Year to date the CFO, H.C. Allen, has purchased +$120k stock.

Items Not Addressed

Dorchester’s participation in wells occurs through a “Minerals Net Profits Interest”. It was established to form a “blocker” of sorts for the UBTI that gets created when the company invests in well programs.

Well participation rights vary by state. In North Dakota you can participate on a well-by-well basis, allowing mineral owners to “cherry pick” the best wells. In Texas you have to participate in all wells drilled in your section.

Risks

Commodity prices, Midland-WTI spreads and interest rates are the most salient