Transcript
Tim Pawul:
Hey guys, this is Tim Pawul from The Minerals and Royalties Authority. I recently sat down with Evan Anderson, cofounder and CEO of Oseberg. During the episode, Evan talks about how he’s helping minerals and royalties companies take a page out of the playbook of hedge funds by structuring and looking at data in a different way in order to buy and sell minerals with a systematic approach. Let’s jump into the episode and hear more of what Evan had to say; Evan, good afternoon, and welcome to the podcast.
Evan anderson:
Hey, thanks for having me, Tim. Good to see you.
Tim:
Yeah, so it’s a new friendship but a fast friendship. I’ve really enjoyed getting to know you these last few months.
Evan:
I came to one of your events; I think I crashed it, and you were the man of the party. So I was like, I’ve got to know that guy: all the good work you’re doing to help others, the scholarship fund, and everything. I just wanted to reach out, and you were kind enough to take my e-mail.
Tim:
Yeah, well, there you go. I’m always the one calling and emailing people, so when it comes the other way, I always make an effort; however, you have a unique background, so let’s start at ground zero. Where did you grow up? Were you around oil and gas? Did you have an oil and gas family or oil and gas ambitions? Then how you got into starting Oseberg? What’s interesting is how you view the world, how you view investing in oil and gas, and how you’ve merged hedge fund quantitative investing with data. Those are pretty much different worlds; at least, they are in oil and gas, and your mission is to bring them together, but over to you for a bit of background.
Evan:
Yeah, absolutely, so born in Montana, of all places, kind of grew up in Oklahoma and spent the summers and winters in Montana. My grandfather had a farm up there and came from a family of academics, so he had no real ties to oil and gas. A family tragedy detoured my plans for law school and got me into the oil and gas industry. So, my first job out of college was helping a privately owned operator in the Anadarko with a bunch of contiguous acreages as far north as Harper and as far south as Stephens County. They operated a lot of wells, and they had a little wireline business, a small little gathering system, and I was learning the business, baptism by fire.
So, that’s how I got into the space, and as I got into the space, I had many friends out on Wall Street doing some pretty crazy things with data; this was back in 2004. When you start thinking about where we were from a technology standpoint in 2004, these big databases Hadoop and AWS cloud were emerging. You were starting to hear from Hal Varian, who was Google’s chief economist, talking about the ubiquity of data and what could be done with it in the consumer space, and that got the wheels turning when it came to oil and gas.
Tim:
OK, great, so walk through your journey leading up to Oseberg. You did some investing before that, right? In 2011, you started Oseberg and developed this thesis around systematic real asset investing. I’d like you to define that because there will be many people who don’t know what the hell you’re talking about systematic investing, but what led you to that path? You did dabble a little bit. Right?
Evan:
Yes, so what I found when I was operating is that there were a lot of datasets that could be created that were unstructured that, given the right technology, you could create structured data sets.
Tim:
And what do you mean by structured and unstructured data sets? That was a new terminology to me.
Evan:
Yeah, right, so I think of structured data in the simplest way: stuff already in Excel or ones and zeros coming from telemetry. A lot of production data is already structured. It may not be standardized and normalized. It may not be aggregated, but it’s structured, so because we filed with a tax commission, we file structured data in a CSV format.
Unstructured data largely involves stuff trapped in natural language documents, word documents, PDFs, TIFs, etc. You’re looking at an image, and there’s nothing structured; there’s no structured information from that image you can do anything with. So it was looking at the world of unstructured data in oil and gas, which goes much broader than oil and gas. It’s really all real assets. Whether you’re looking at renewable development plans, or you’re looking at real estate, or you’re looking at mineral royalties working interest, there’s a lot of really valuable information that’s in legal contracts and regulatory filings. These PDFs and TIFs contain just some very valuable information that is not in a useful format.
Tim:
And you just had this epiphany as an operator, you had all these documents, and you’re like, if I were to invest, this would be interesting. I mean, how did you come from point A to point B?
Evan:
On the courthouse records side, you file oil and gas leases, assignments, mineral deeds, and all sorts of documents related to real property. That was a starting point. I would spend my time at the OCU law library in Oklahoma City reading the published oil and gas law handbook that Eugene Koontz put out, and Owen Anderson had a supplement in the back. I wanted to understand how these agreements I wanted to understand implied and explicit covenants, the language used in those covenants, and what’s the material value of that language. How does it actually impact an asset? I started thinking about what if you could structure that information rather than serendipitously stumble upon it and search it and derive some analytics and insight from that structured information. Then being an operator, you’re constantly filing various regulatory filings in the case of Oklahoma with the Corporation Commission or in the case of Texas with the Railroad Commission or the OCD in New Mexico, and a lot of the regulatory filings, whether it’s a nonstandard proration unit in New Mexico or whether it’s in increased density in Oklahoma or a P8 in Texas a lot of these regulatory filings were ultimately PDFs or paper that was submitted to the various regulatory agencies.
So in the early innings, before I started Oseberg, I started manually structuring that information to derive some intelligence, so I was looking at the Arkoma when it was first kind of developing, and I was looking at spacing activity, and I was looking at the options offered in a pooling; option one, option two, option three and it wasn’t making a whole lot of sense why the first option was I don’t know an eighth royalty and $500 lease bonus. Option two was a 3/16 and $250 lease bonus. Why did somebody determine that a 16th of an acre or a 16th royalty was worth $250? Because when you backed into the AFE of the well, which was also in the public domain, it could be a $9 million well. That didn’t make any sense, so you need to structure that information to perform that kind of analysis, and then also those poolings were indicative of future development, and there are all sorts of dates in those poolings. I wanted to start looking at well, if a pooling is filed, what’s the probability that it’s actually going to get drilled, and how long does it take before it starts producing, and you can start looking at some behaviors of the operators and get a sense of how long before that acreage actually starts producing. Some of it was on the courthouse record side, where I started collecting some of these ideas just by reading about some of the documents that I was working with, and some of it also had to do with operating and being on that side of the business and having to file stuff with regulatory agencies.
Tim:
So some light bedtime reading you’re tinkering around, you start doing it manually, and do you start a little pilot investment fund using some of this data would you do with it.
Evan:
I wish it were an investment fund. No, I just started putting acreage together. I just started doing deals, and I could move a lot faster and get a little ahead of where the puck was moving due to creating those data sets. Really what I wanted to do, I was in my early 20s, and what I wanted to do was kind of what you had described, build a systematic approach to underwriting and then execute against the strategy to acquire minerals, royalty, or working interest. The problem is that there’s a lot that data can do to accelerate the evaluation, the deal flow, and many aspects of that workflow, but it’s still as much art as it is science. You still need that qualitative background. You still need an exploration geologist or reservoir engineer to look at the data and draw upon their qualitative experience to come to some conclusions. I wasn’t in a position in my early 20s to put together a management team with that experience because most of my industry peers were just starting at the independents and majors.
Tim:
Got it, OK, so talk a little bit then about Oseberg and when you started it and kind of just the early days, and then we’ll kind of loop back to again the thesis and the mission statement of getting all gas companies to invest systematically.
Evan:
Yeah, absolutely, so it was really ugly. The reality is that if you’re trying to build a data and technology company in Oklahoma, Texas, or Arkansas, it’s a tough slog. You know that we are real asset communities, so there are not a lot of founders to tell you how to do this. I thought that raising money for a data and technology company, which I decided to do, would be easy. It’d be like putting together a deal to drill well. I found that after four months of nonstop pounding the pavement, I wasn’t getting any traction, and my father-in-law at the time looked at me and said you need to get a job; if this idea were good, somebody would have funded you. So honestly, there was a gentleman in Oklahoma City that I had met his son, and he had heard that I was out trying to raise some money. He was kind enough to give me what he thought was going to be an hour, and after what I’m sure he would describe as three hours of blunt force trauma, I turned to him and, in desperation and said, is this a good idea, I mean just tell me if I should move on, and he said it’s a good idea and I’ll fund it. I’m sure he was looking for a tax write-off at the time, so he wrote me a $250,000 seed check.
I’m a non-technical founder, and that’s the kiss of death for a technology founder, and to say that I lit that cash on fire and I didn’t have a proof of concept and all the problems with building software when you don’t know how to build software. Where you don’t have a locked-down roadmap and requirements, I went back to him after I was coming down to my last dollar, and I’ll never forget we met for breakfast, and I think he knew it was coming and I said, look this time I know what I’m doing, and I need another $250,000. He told me that he would do it under a convertible note, but he told me to take my time to find whoever I needed to get this thing in the market, and I found this guy that had some ties to Oklahoma; he was at EDS, and he was looking to get back to Oklahoma. he had some pretty strong data engineering skills, and I said look if you get this thing to market in six months with me, I’ll give you some points in the business and make it worth your while, and we worked from 6:00 AM to 9:00 PM for six months got that sucker to market and did $500,000 of recurring revenue in the first couple of months.
So that was the spark, that was the beginning. Look, the company was built off of a thesis that real asset investing would become systematic, but you need a lot more structured data to do it from the whole workflow, from underwriting to deal sourcing to putting capital to work, to how you manage an internal buying team, or how you manage assets. I needed to create many structured data sets to do that, so at first, we were bringing data to market, but we realized that the problem was much bigger than we had ever anticipated. So at first, it was just about creating data, and then it quickly became about building infrastructure and technology to structure data at scale. When you look at the real asset market, whether it’s real estate or it’s oil and gas, or it’s shipping and logistics, there are a lot of markets, there are a lot of companies that desperately need datasets created, and it’s an expensive endeavor without that technology.
Tim:
Give me a 30,000-foot view of the data space and energy as you see it. What’s typically focused on, the types of data, how they go about it, the problems they’re trying to solve, and where you saw a lane to go in and disrupt where people didn’t seem to be focusing on it right or have been able to solve.
Evan:
Yeah, good question. I don’t know the numbers right now, but let’s say that upstream CapEx budgets are $150 billion a year. The big chunk of that is in drilling & completions, so when you look at what most oil and gas companies think of as data, it makes sense that many of the datasets are around the drill bit, drilling & completing. We’re talking about production data. We’re talking about well data. We’re talking about seismic. We’re talking about well logs, and that’s where many of the technology incumbents have focused. When I came to market, I knew that it was going to be difficult to differentiate ourselves, I had some ideas, and some of those ideas we executed, but where I wanted to focus was what I just thought was going to be a much bigger problem that would touch on a lot more industries.
That is first and foremost in the courthouse records. The problems that you have with courthouse records are collated and uncollated images, you have poor image quality in terms of the scanning that’s been done, or if it’s a photograph, you have duplicates, you have incomplete scans, you have strike-throughs, you have addendums, you have handwriting. There’s no metadata. There’s no data around the images. The indexes in courthouse records are a mess. New Mexico, for example, courthouses have zero metadata, and what that means is that you’re flying blind; you don’t know how to locate the image that you want quickly. The oil and gas lease, marriage license, or whatever else you’ll find in the courthouses, the mechanics lien. There’s no easy way to locate the oil and gas leases or even filter them beyond that, who the lessor or lessee is or the terms you’re looking for. So I focused first and foremost on the courthouse record space because many industries rely on courthouse records and struggle with that problem.
The other area that I looked at was the regulatory space, which I describe as it sits between leasing and drilling & completing, everything in the middle. It’s the process by which companies make filings to drill and produce. There are many different filings in the public domain, but getting your hands on them at scale isn’t easy. You still have the same problems you encountered with the courthouse records. In some cases, you get a PDF file that is 70 pages long, and the information you want is on page 55, so you can’t manually do that when you’re looking through hundreds of thousands of documents. Even if you try to offshore, it doesn’t work at scale. So the regulatory environment had a similar set of problems, and there was a lot of high-value data within those documents, and that’s what I was chasing. Once you aggregate all those documents, once you write the code and do the preprocessing so the image quality is improved and you can OCR those documents, which means taking characters on a page and making them structured effectively, then you can start developing applications that can bring those characters and put them into a structured database so that you can actually do something with that information and derive insight from that information. So that’s where we’ve focused, I mean, there are other areas of data within oil and gas, but that’s where we focused.
Tim:
Returning to the question of investing in real assets systematically, we’ll define that again. We’ll go into where you think we are and, in the baseball analogy, what inning we’re in with adopting that in the old gas space since you’ve been on this endeavor for a little over a decade.
Evan:
Systematic investing, or quantitative trading, uses extensive data sets to find patterns. Historically, it’s been done in finance, and over time data has primarily become commoditized in finance. You’re rarely able to find information asymmetry, a piece of information your competitors don’t have, so as a result, speed is the edge in systematic trading. It’s different when you look at real assets, oil and gas, wind, and timber because much of the valuable information is in these unstructured documents. It is taking a systematic approach to deal valuation, underwriting. It is taking a systematic approach to deal sourcing, trying to determine who you will buy from. It’s handling data and determining the probability of something actually getting developed and produced. It is how you run your team in terms of executing and deploying that capital. Something as simple as using a CRM is rare in our space.
One of the groups that I spend some time following is Two Sigma. Just by way of example, Two Sigma brought in Tom Hill, who was formerly with Blackstone, and they hired the guy responsible for managing WeWork’s real estate investment platform. Two Sigma is taking the approach where they can use a lot of structured data to drive a competitive advantage in identifying opportunities in the real estate market. Now Two Sigma is a very different animal than your typical management team that you find in upstream oil and gas, these are the type of folks that you go into their kitchen, and there’s a sign that’s sitting above their sink that says in God we trust all else bring data. In everything they do, they look through the lens of data and technology. It’s the first thing they think of is where are the competitive advantages and data and technology. That’s very different from what we see in the real asset community, where data and technology are more of a service to what you do than your core competitive advantage. It’s not the first five pages of your LP pitch deck; it’s page 45 out of 50.
Tim:
Yeah, that reminds me. In my world, I do a lot of content. There’s a guy I like called Gary Vaynerchuk, and he always says in today’s business, you’re a content or a media business first. Then second, you’re an insert, exploration company, minerals company, real estate company, etcetera. He said that has to be the number one, which makes me think of this as saying every company in your eyes that does real asset vesting should be a technology company first.
Evan:
Absolutely, when I see downturns, data analytics teams are some of the first to be let go. When I look at management teams that are getting funded by private equity, it’s rarely the case that somebody on the management team is a CTO or CDO, chief data officer. Those skill sets are not on the management team or the founding management team. When you look at the two and twenty model, that 2% of G&A will not buy you a lot of data in tech. If I showed a systematic fund how to drive alpha with data and tech, they’d write a $10 million check on the front end. It’s just a very different animal and very different culture, and you’re exactly right, but what’s interesting about that is our industry has always relied on information asymmetry; it’s always been about information, and it’s interesting that isn’t the central theme by which we pivot around.
Tim:
Yeah, sorry I cut you off you. You were talking about Two Sigma. Were there any wrap-ups or following thoughts to that story you were telling
Evan:
I think they represent where we will see our industry go, and when I think about what drives change, it’s usually two things. It’s somebody crushing everybody else, outsized returns, somebody’s insanely successful, and then you’re going to see follow-on, or it’s absolutely got to do it, or else we’re going to die necessity. Even then, we’ve seen negative $30 oil, and companies are still spending millions of dollars printing checks and spending money on paper and ink. I mean millions of dollars in checks. If you look at the process by which the division orders are sent out and received and input, it’s archaic, and you don’t just see that in oil and gas. In shipping MERSK, all their shipping logistics information is on paper, and these are enormous industries, especially in the real asset space. When I look at what companies like Two Sigma are doing where they’re using predictive tools to forecast demand and supply when they’re running more of a systematic end-to-end approach to deploying capital and executing whatever strategy they may have, I think that with a great deal of success, there’s going to be a lot of follow on and that’s going to bleed into other sectors.
Tim:
Super interesting, where my mind plays devil’s advocate or conspiracy theory, the archaic-ness and inefficiency of how division orders are filed, at least on the royalty side. Right, how arduous it is to get into pay status or the different things that can happen to get you out of pay status, and I look at the hard cost of that. Also, operating cash flow is the name of the game. It’s the lifeline of the business. So if you can delay paying a royalty owner for months and even years on end when you’re at the scale of a large independent, that’s just crumbs. It doesn’t really matter, but if you’re a smaller business and you defer millions of dollars and royalty checks until higher prices come in, that’s interesting. Maybe you’re not incentivized to make it more efficient.
Evan:
No, that’s a super-valid point. It’s like when people complain about the timing of production data that are reported into the state and made available for those with a working interest in every well. What’s their incentive to lobby for the state to become more efficient? It doesn’t bother them because they’ve got stuff coming from check stubs, they’ve got stuff coming from telemetry, they’ve got that production data, but for everybody else, we’re waiting three months before that data hits the state website, we’re looking in the rearview mirror we’re not looking at the present so that’s certainly a valid point. Some of the inefficiencies are also competitive advantages for these companies, without a doubt. I don’t know if that will last very long.
It’s interesting. I was meeting with a company, and a gentleman observed that we wouldn’t see any actual leasing activity in the next five years and that leasing data would largely be irrelevant. His thesis was that a lot of the core rock would be HBP. The next five years was his time frame, and I stepped back and thought, if that’s the case, let’s assume that we’re not going to see a lot of turnover and leasing activity in the core of the core that means that mineral owners aren’t going to see these new liquidity events of bonus payments and things along those lines.
So what are you going to do if you’re a mineral owner and you’re not going to see another big lease bonus check? You’re going to start looking at your 30-page lease agreement, and you’re going to start looking at all the provisions that are not tracked by companies. They have no visibility into what encumbrances they agreed to, and there are a lot of breach which means that title litigation funds are going to be coming along in this space too. Because technology is now at a place where we can cut through 88,000 leases in a heartbeat, structure all the continuous development clauses then make a parent-child relationship, and give a lot of insight to somebody that wants to be proactive about enforcing their lease agreement.
Tim:
That’s a super interesting insight. People always ask me, Tim, how much opportunity is left in the middle space? It’s getting so competitive. Where is the low-hanging fruit? Where’s the opportunity side going forward? It’s a question I sometimes ask, at what point do we run out of minerals to buy? But what ends up happening is opportunity recreates itself, and you get these windows of opportunity. That could be technology-driven like you’re talking about; it could be Black Swan events.
So the example I always give is through 2020, when the drilling moratorium came on from Biden in New Mexico when oil went negative when COVID was in full force. New Mexico was off the table for the majority of minerals funds, and a handful of folks leaned in and got very aggressive in buying some of the best rock in the world, and they were able to get an excellent cost basis. One of those individuals is Nick Verrell at Wing. He built a world-class asset in 12 months and sold it for an incredible return before and after that; New Mexico was ultra-competitive and pricey, but that was a window of opportunity that was created, and you can’t predict these things.
That’s what opportunity looks like in the mineral space going forward. Technology and transparency make the secondary market more liquid; you start to have these mineral interests listed freely so that retail investors can trade them on an exchange. There are so many places to go, and I think that’s what’s exciting about it. Then just the day-to-day business as is, there’s still so much opportunity and information asymmetry, and that’s the world you’re looking at today. Right?
Evan:
The way to answer that question depends on the structure of the financing coming behind that question. There’s no doubt that mortality rates are significantly higher than they have been because the baby boomers are aging out. So that means that we’re seeing a lot of probated estates. We’re seeing a lot of what was one royalty check becoming five becoming ten, and that’s happening across the board. So the market is becoming much more fragmented. Whether or not you’re looking at one net mineral acre or 1000 net mineral acres, the transaction costs are relatively the same regarding the required title. That makes things incredibly challenging, and data and technology have to come in and fill that role because something has to change in terms of how the title is being performed for that to be economical. You will have to change how you source and buy minerals as they become more fragmented.
Tim:
So as a follow-on everything we’re talking about, it always struck me that you’re like, Tim, what frustrates me. I have difficulty getting clients or prospective clients to think differently, and everyone talks about production and lease data. You always tell me, Tim, it’s $#@&ing data, we created these artificial silos, and that’s actually creating so much tunnel vision and so much inefficiency, so comment on that and how you see the world and how one of your goals is to break down those silos.
Evan:
Absolutely. I make the offer to any management team I meet with to consider leveraging data to buy assets. Whether they’re a customer of mine or not, that is a passion of mine. I believe in a pure meritocracy. Some of the incumbents that we compete with have done a phenomenal job; there are some things that I don’t even endeavor to contend with. Some things we have done are unique and differentiated; I can hang my hat there. But if you’re open to learning about leveraging data, I’m open to visiting with you. So your point about some of the lexicon that’s been developed, who decided that regulatory documents are land datasets.
We have these models in our head that this is land data. This is engineering data. This is geology data, which gets under my skin because it’s about what’s the problem, what you are trying to solve. What are you trying to do? Are you trying to drill better wells? Are you trying to accelerate the deployment of capital in an area? Are you trying to drive MOIC and make better bets? What is the problem you’re trying to solve? Start there and then talk about the world of information and what’s available to give you a competitive edge in achieving that goal. I can give you several examples where what someone would describe as a land data set I would fundamentally describe as an exploration geology data set because when you step back and ask why some of these regulatory filings exist? What purpose do they serve? It directly ties back into reserves or to exploration, and so the cultural lexicon and fixed mindsets that exist in our space are something that I’m passionate about breaking down.
Tim:
Let’s have some fun here, entertain me in the audience. We’ve been talking high level now but walk through some theoretical examples. Oseberg has created these structured data sets. This is a minerals podcast, so how have you seen it applied to minerals? How have you thought about applying it? How have you seen clients and management teams you’ve spoken to apply it? Let’s just get some different examples that are kind of fun and topical.
Evan:
Absolutely, ok, so am I a mineral buyer, or am I a mineral owner?
Tim:
A mineral buyer for the sake of this podcast.
Evan:
So if I’m a mineral buyer…
Tim:
By the way, if you’re a mineral owner, as in a fund or portfolio that owns minerals, go down that rabbit hole as well.
Evan:
If I’m a mineral buyer, the first thing I need to do is figure out what I’m going to pay for minerals, and there are a lot of great products out there to help you with underwriting. But one of the questions I ask teams is what insight you have around PUDs and the development of white space. Because it’s tough to buy PDP right now, there’s a lot of capital chasing PDP, and so you got to try kind of to shake free from that if you’re going to underwrite PUDs, I’m sure you have some valuation methodology, but do you have any insight into the timing of development.
So let’s take Texas, where people love to tell me there’s left there’s a less robust regulatory environment in Texas. There are more regulatory documents in Texas than there isn’t any other state. However, there are more regulatory documents in Texas than in New Mexico, Oklahoma, or Wyoming. In Texas, if I’m looking and I’m trying to predict development, I’m going to go to the courthouse records, and I’m going to look at the lease provisions, the continuous development clauses because if companies don’t drill out and earn out their acreage they’re going to have a reserve write-down.
So that will be an informing variable regarding well planning for a company. It’s not uncommon that I talk to some companies that don’t even have that visibility internally about the continuous development clauses and those obligations. But increasingly, I’m meeting with many companies that do or are working to, and that needs to help inform their well planning. They can’t afford to breach an essential oil and gas lease and then have to write-down a massive chunk of acreage. So if I’m a mineral buyer and I want to step out a little bit of what’s already producing, I might be looking at those continuous development clauses to help inform me as to where development might be taking place next, and that allows me to buy at a better PV value and that’s what’s going to help drive your MOIC.
Tim:
Now, what’s really interesting, and I’m sure there are some folks out there, a lot of intelligent people in the space that have looked and executed on that as one of the variables. However, I have spoken to many folks, and I’ve never heard that before. Going to your point on structured data, people are chasing everything structured. A company I work very closely with SourcEnergy uses satellite imagery to track development, but they’re making structured data sets out of aerial visuals, pad development, and infrastructure development. What they do is starting to become somewhat of a norm for people to use as criteria, and this is all publicly available. Keep going; this is super interesting.
Evan:
If I move to Oklahoma or New Mexico, leases have already been taken. We’re talking about where’s development going. I’m going to start looking at the regulatory process because it’s a linear process; it moves from left to right. Meaning you can’t file a pooling before you have spaced a formation. So you have to file a spacing first, then a pooling as I move from a lease to a spacing to a pooling, and I’m oversimplifying here. I’m getting closer to actually drilling a well. If I wanted to benchmark operators, I would be looking at those regulatory filings. I’d be keeping track by formation by the operator, the probability that they fall through, and how many days, on average, they actually drill that well when they file that regulatory filing. Then I would compare that to their quarterly calls to see how many wells they said they would drill and how many wells they actually drilled. Those are all input variables that should go into underwriting when looking at PUDs and white space. You can delineate the trend by looking at the wells and the leasing activity. Now you need to look within that lasso and start looking at some regulatory information. Maybe it’s not a perfect science, but it should help inform your view. so there are still some things that could be done on the underwriting side because minerals aren’t worth much if they’re not producing. So there’s a lot of data that has been historically unstructured that we’ve worked to structure. There are still other data sets we’d love to structure that can help inform some of that underwriting.
Now if I’m a mineral buyer, we need to figure out who these people are that we need to buy from. I mean light reading. I like to read state statutes. So if you read the state statutes, they do a great job of spelling out in painful detail what you must file to operate in their state. It’s easier than a driver’s license exam. It just spells it out. What you’ll find is, no surprise, that anytime you’re going to impact the value of anybody’s mineral state, they’re going to have to get notified; you can’t just do that; you can’t just change the spacing and go on your merry way you have to inform them, and you may not even have to notify them you may also have to notify all the operators in the area because it may impact their wells. Well, that notification process is pretty nice if I need a source of information on who to buy from. Now some people will say, “Well yeah, Evan, there are a lot of problems with this data set that need to be addressed. It’s not perfect, but it’s a great starting point with a good list of names and addresses. In some states, they tell you who’s a working interest, a royalty interest owner, and an overriding royalty interest owner. In some states, they give you a lot of details. Some states tell you what percentage they own, others don’t, others mix mineral interest owners’ stakes with working interest owners, and you don’t have an easy way to determine who’s who. Well, if we compare that list to our lease database, we can start getting some matches. We can begin to filter out who’s working interest and whose minerals. Then I’m going to have somebody say well, Evan, but they don’t tell me how many net mineral acres they own, and I don’t want to be calling on one acre. Well, let’s just put our thinking caps on here. Companies building out positions also don’t want to lease one acre at a time, so go back to the lease data and look at who was first leased in the trend because typically…
Tim:
And they’re going to be the larger positions, of course. Makes sense.
Evan:
And it’s not uncommon in some places where you see significant development in oil and gas that many surface estate owns the mineral state. So if there are lots of regulatory filings in an area, they’re going to be notified and notified and notified and notified and notified. So you can start looking at the frequency of somebody being listed as a respondent. That gives you some information about them more likely being a more significant mineral owner. It’s not guaranteed, but in a world where you have a choice where you can go spend a fortune sending people to the courthouse and manually read this information and take forever versus immediately having a spreadsheet and having a data analyst that has about as many skills as a college intern to massage some of this data, you can get a lot of intelligence relatively quickly.
Tim:
You had some math behind this earlier in the episode. You said a lot of this can’t be offshored, and what you’re talking about is if you were to throw tons of cheap labor, given it’s a different currency and a different country, you still can’t go through this. Put the math behind that because I thought if you had someone brilliant like yourself who can connect the dots and know where to look if they want to buy acreage in a specific area, they could piece this together, but how can this be done outside of a manual process? There are too many moving parts. Talk to me about the machine learning aspect and the equivalent of human hours. It’s a stupid number, the equivalent of dollars spent on human hours, you told me earlier, and without machine learning, none of this is even remotely possible.
Evan:
First of all, I want to qualify that I like technology, but it’s really a scalpel. You need to know what technology solutions to use for what problems. Everybody loves to talk about artificial intelligence and machine learning, but the number of people talking about it probably don’t fully understand it.
Tim:
Kind of like myself…
Evan:
There are technologies that are a couple of steps removed from machine learning that bend the arc, something that would take you 30 minutes to an hour shortened down to minutes. So you can get huge gains without entering the machine learning world. I’ll give you an example. We wanted to create a database of names and addresses from some respondents we pulled from New Mexico regulatory filings. We also wanted to structure some courthouse records lessee names and addresses and mineral deeds names and addresses. We were looking at about 24 million names and addresses from these documents. It was 17 million documents and 24 million names and addresses. We’ve worked with extensive teams offshore for almost a decade now. We’ve done a lot of different projects, and we have a pretty good feel for what things cost and how long things take. Analysts in India would charge you for a name, an address, a document type, some dates, let’s call it ten-column headers. They would charge you about $0.40 for a name and address. You’re looking at 24 million names and addresses and spending over $9 million, and people choke when they hear that, but that’s not the worst. Let’s say you had $9 million to burn, and you’re going to do that. You’d never get through it because one person can do about 200 entries a day, so if you had 400 bodies, it would still take you two years to do it. I don’t know if you’ve ever tried to manage 400 people on a project, but it’s not a trivial effort, and so is the reality…
Tim:
And then the documents replenish. You’re constantly falling behind, right?
Evan:
Exactly, and some of the things that technology enables you to do is not spend $9 million but spend half a million and have something done in three months. That’s the difference which is insane and more achievable. Then rather than spending your $9 million creating the data, you can spend a couple of hundred grand standardizing and normalizing, so when I see your name with 36 variations because people don’t know how to spell, and when I see that you have 48 different PO boxes, when I see that some documents you sign with your wife other documents it’s just you. I can spend some money standardizing and normalizing, so I get one name for Tim Pawul. Then I can start cross-listing that when I begin standardizing and normalizing data with other data sets to learn a lot about Tim.
Tim:
Super interesting. Do you have any other examples as a mineral buyer to walk through?
Evan:
Yeah. We discussed using the information to predict development; again, it depends on where you’re looking at and what basin you’re in. It depends on if you’re buying working interests or minerals or MRIs. I go back to forget about the data for a second let’s start with what’s the problem. What are you trying to do? I need a lower cost to identify who owns what. That’s your problem statement, or I’ve got a lot of money, and I’m having a heck of a time getting it deployed. I need to accelerate the deployment of that capital, and I’m also fighting over PDP, and everything’s too expensive. I need to step out and buy some PUDs and white space, but I don’t want to buy stuff that will never get drilled. So whatever your problem is, you start there, and then you start looking at data. We talked about the probability. We talked about figuring out who to buy from downstream. From there, you’ve got a process problem.
I see this a lot, an individual has a spreadsheet in front of them, and they’re making phone calls, and that’s it. I love talking to management teams and asking them who’s your best buyer and it’s interesting to hear them say why Tim is their best buyer because they rarely have the KPIs to support why Tim is the best buyer. What are you tracking? Why are you running that process? What technology tools are you using to run that process? Tim collects great information when he calls and connects with the mineral owner. Do you guys have a disciplined approach to managing that information so that you can leverage it? Because Tim will tell you who his cousins are and his auntie is, he will tell you whether or not he buys in the name of just him or him and his wife. It’s just a ton of valuable information that’s not getting put in anything helpful to you. We call that data exhaust, which goes back into the feedback loop of underwriting, deal sourcing, figuring out who to buy from, and executing, and that’s the systematic acquisition of real assets. It’s not just an oil and gas problem.
Tim:
You mentioned being a mineral owner. Are there any different ways to look at it?
Evan:
Yeah. Remember, in my early 20s, I was looking at the Arkoma and telling you about these pooling bonuses. If they’re going to force you to lease through a pooling in Oklahoma, they have to testify as to what the market rate is in the adjacent sections that they’re pooling. What are they offered and paid for? They have to give the mineral owner three choices. So as I was describing earlier, you have option one, which is an eighth and $500 an acre; option two, which is three-sixteenths and $250 (less bonus, more royalty); and option 3, zero bonus and 25% royalty. Has anybody ever said I will take 50% of my acreage and choose A and 50% choose B and hedge it? How do they graph that? I will take some cash and put it in my pocket but also blend the royalty to get a higher royalty if the well hits. Also, how do I know what it will take for this well to be MPV zero and pay back? The AFE data is in the public domain, and everyone will say AFEs are never accurate, but are they 50% inaccurate, or are they 15%, give or take? That’s an input. In addition to that, what is fair for my minerals and that pooling data is a good indication of what your minerals are being traded for. So as a mineral owner, a ton of data is available to make them more innovative sellers.
Tim:
Super interesting because it feels like shooting in the dark trying to get a feel for market pricing, and there are two layers of it. But your data exhaust example is super interesting because that’s my world. I collect so much valuable intel from conversations…
Evan:
What do you do with it? Is it in anything you can use?
Tim:
It’s in a spreadsheet, and I do better than most to extract value from it. But yeah, it could be way more sophisticated.
Evan:
Yeah. I talked to somebody about this earlier today, and I have a lot of empathy for the space because you have these subject matter experts who know how to value minerals or working interests. They know how to transact, the title, and how to close the deal. But they probably have a friend in IT, they don’t know where to go, and they don’t know how to endeavor to build something themselves. You see this a lot. It’s like me in the early innings, at the beginning of our podcast, where I’m lighting $250,000 on fire because I don’t know how to build tech at that point. I hired my sister’s friend, a project manager at Samsung. Everybody in technology is laughing because a project manager differs significantly from a product manager. It took me a while to learn the difference. If you want to do something with data, where do you start? Who do you trust? You talk to the person closest to you who knows anything about technology. I can’t tell you how many phone calls I get, “I’ve got an app idea!” I’m like, “That’s awesome. I don’t know anything about app development.”
Tim:
What is that British sitcom? It’s a comedy. It’s very old. I think it’s called the IT Guys or something. it’s very dry and sarcastic. It’s very British. These guys are sitting in this office, and it’s a bunch of calls, and they’re not paying attention. “Have you tried turning it off and on? Did it work? OK, great.” So for me, I have my best buddies from Rice, where I went to college, and they’re developers but anything IT related I just…
Evan:
What kind of developers are they? Are they application developers? Are they data…
Tim:
I have zero idea
Evan:
Totally totally, but you’re going to call. That’s your first call, right?
Tim:
That’s my first call.
Evan:
“Hey, I’ve got these heart palpitations; my buddy’s a dentist. He’s in healthcare. I will give him a shout and say can you listen to this? what’s going on?” He is going to say, wrong anatomy, dude. I’m empathetic to it because I was there at one point. I mean, we’ve spent a fortune building technology and hired and fired tons of engineers. You learn, but it’s an expensive and challenging journey to do it yourself. I meet some great management teams endeavoring to do some cool stuff, and they’ve probably been burned. They’ve probably tried to do something around data and technology. They’ve probably spent some money and didn’t get anywhere. One thing I never forget when I talk to management teams is that these guys have been around for a long time, making a lot of money without me. So who am I to tell them that they need to be doing anything different? but when I hear, “We’re trying to deploy capital in the Delaware, we’re struggling.” I lean in, “Well, are you ready to change? Are you ready to try something different?” when I hear, “We’re fighting for PDP. we can’t move capital.” I lean in. “Are you ready to change? Do you want to try something different? Does it hurt enough? I’m here to help whenever you are ready.”
Tim:
Your analogy is spot on. Someone either crushes it, and everyone goes holy $#@& what is that mineral company doing? Why are their IRRs through the roof? Word gets around. Or the opposite. It’s a knife fight. I can’t tell you how many times I heard that. It’s a knife fight on the ground. So it’s enough pain, and that pain that might not be COVID, deal flows locked up, oils negative, we’re laying people off., it may not be that pain. It may be we can’t deploy enough capital, we may get the plug pulled, or we’re not going to get enough capital deployed and get the portfolio where we want, where we hit our pref rate and make money. Pain can mean a lot of different things to different people. But if the pain is that I can’t deploy enough capital, or it’s getting too competitive, you say, why don’t you think differently? Let’s lean in. I think you’re going to have a lot of meetings. I know you’ve succeeded with the folks who’ve worked with you. So let’s wrap up. I knew this would be fun, so thanks for coming on. I always enjoy talking to you.
Talk a little bit more about your wheelhouse. The basins you’re active in and who you love to talk to in the mineral space. Make it very simple if you fall into these buckets, let’s get together. You’re based in Dallas but can access anyone via Zoom or travel. Handing it over to you to wrap this high-level conversation up and synthesize it into a follow-up if people find you intriguing and want to engage further.
Evan:
Yeah, I’ll open it up by saying you don’t have to do business with me to reach out. I want to talk to you if you are open to doing something different. I want to help you. I will help you hire engineers. I’ve got a long list of folks to connect you with. I will help you use some of our incumbent’s data differently. This is truly a passion of mine.
So we’ve sunk an enormous amount of money into building the technology and infrastructure to scale data from unstructured documents. As a result, we’re limited in geography. Our focus is New Mexico, Oklahoma, and the Permian in Texas. We’ve got your traditional datasets, well datasets, and production data sets. There are differentiating factors between our data sets and the incumbents, but where we excel is in the courthouse and regulatory spaces. Some of these incumbents will tell you they have regulatory data, and that’s true at a 100,000-foot value level. They do have regulatory data, but the difference is they don’t have the same data set that we have, and they don’t have the same vision for how to leverage that data set. They certainly don’t have the support team to help your team to utilize the data in the ways we’ve been discussing today. That’s what differentiates us. We love to talk to anybody transacting around minerals, working interests, upstream development, and midstream companies. I just worked with a gas marketing company that wanted to grow in the Delaware. They needed to determine who the working interests owners were they could contact to market their gas. That was a layup for us—business and commercial development for midstream marketing. We talk to people in land, we speak to people in engineering, we talk to people in geology and legal those are all the different disciplines we’ve worked with in the past.
Tim, we’ve discussed the cultural challenges in adopting data and technology. There are real financial barriers for some of these management teams. When you only have 2% of whatever’s committed, you don’t have the capital to invest in data and technology. But people forget to back that software data investment to a hard ROI. When I talk to companies and price something out for them, it could be as few as 24 net mineral acres to put you in the money on a piece of software. If you can’t find 24 net mineral acres out of the ~24 million names and addresses in our dataset, then I don’t know what to say about that…
Tim:
You should get into a different business.
Evan:
I didn’t want to say that, but probably. We do lots of different projects. We structure information for companies, their proprietary documents. we license, and sell data. we license and sell applications to interact with our data. That’s our business, and it’s been a blast talking to you.
Tim:
Awesome, well, thanks again for coming on. I think everyone’s going to really enjoy this, and I’m going to enjoy cutting it up and repurposing it on LinkedIn as well as some good sound bites here, so looking forward to seeing you in person soon appreciate it again; and have a good evening.
Evan:
You bet.