AI-Assisted Due Diligence for PUD Leasehold
6
Minutes
Written by
Brandon Dodds
The Challenge
Dudley was approached by a client who tasked us with evaluating a significant leasehold position they would be acquiring. They also wanted us to organize the data for the producing units into packets that they could utilize post-trade.
Per our service agreement with the client, we had just 22 days to review and submit defects.
The client intended to utilize an AI platform called ThoughtTrace to review the leases included in the trade and then provide that data to us. They allocated value to the wells in this trade, of which the majority were proven to be undeveloped reserves (PUD wells).
We decided it would be feasible to review the wells representing 80% of the trade value within the allocated time.
The client hosted some pre-PSA items that were provided to their own virtual data room (VDR). However, the seller was delayed in providing their own VDR, which created a setback. Furthermore, the seller’s VDR was an open-ended search format rather than formatted into unit or well folders. The search engine was powerful, but it was a learning curve for our team to determine a research routine that we felt would cover all pertinent items.
We had our work cut out for us.
Our Strategic Approach
First, we took data from the provided trade spreadsheets to create a master spreadsheet. We sent a landman to do an initial lien and encumbrance check at the physical courthouses involved in the trade and had another begin production gap analysis on the wells.
While doing the production gap analysis, we tied the wells to their respective units.
By doing so, we determined that reviewing the top 13 units would cover 80% of the trade value.
Consecutively, we had a landman begin a review of material contracts involved in the trade. That data, as well as the production data, were merged into the master sheet. The spreadsheet also tracked the assigned landman and various fields to verify our ownership findings per well. We set equations to calculate the monetary value of any variance and to identify if the defect in question would trigger the defect threshold set by the PSA.
We added a summary tab to the spreadsheet that would pull pivot table data so we could show live results of the value being reviewed/completed and information on the count/value of defects identified.
We also created a due diligence worksheet (DDW) for each unit, which would compile all pertinent information. That form would include tabs for well info, unit WI/NRI calculations, title opinions, lease review, land contracts, production reports, maps, surveys, and a summary.
We used a task management service to delegate assignments and post team messages and oft-cited documents. Some important items we posted were templates, payout status lists, REV decks, JIB decks, and division of interest statements. We also enabled a chat room within that service where the team could discuss issues amongst themselves. That became invaluable to expediting the learning curve of the virtual data room and ironing out frequent leasehold assignment situations with the seller.
We took as much information as we could download from the client’s VDR and hosted it in an FTP site with optical character recognition (OCR). Doing so allowed us to scan documents and images into machine-readable, editable text and quickly search for pertinent documents.
We proceeded to pull data for a starting point, typically a title opinion. We would then look for any documents past that date that would affect ownership and calculate a verified working interest (WI) and net revenue interest (NRI) value. We would reference the payout status reports to ensure we knew the well's phase. With this trade, there were multiple APO values to track. We also compared their purported ownership to what we found on the division of interest documents to make sure nothing appeared off.
Setbacks & Solutions
Our material contract review focused on some situations where the seller should only have inherited wells if they were timely drilled. Furthermore, some prior agreements would allow other majors to come back in on future development. This was a concern since our client was mostly obtaining PUD properties.
Additionally, the AI platform our client used to review the leases did not provide the exact data they needed.
Unfortunately, we were two weeks into the project by the time the data was organized enough to upload. We discovered that the way the information was formatted when uploaded caused output issues, rendering the data unusable.
It did OCR the full clauses fairly accurately. However, it was unable to consolidate the clauses into common categories. This resulted in a huge redundancy of columns that were essentially handling the same information.
The client then asked us to do the lease review ourselves, since the results caused validity concerns. Since only a few weeks remained at that point, we limited our review to high-priority clauses, only as to leases that were included in the unit we were reviewing in detail.
We pulled this data into the master spreadsheet and the respective due diligence worksheets. We also compared the leases we found on the title opinions and unit declarations to make sure that the seller’s exhibit included all leases of value.
Key Outcomes
Ultimately, we were able to exceed estimates and compile due diligence worksheets for 16 units, which covered 89.71% of the trade value. Despite some setbacks, as a team, we found ways to work through these issues and provide the client with the information they needed.
The due diligence worksheets have proven valuable to the client. They could quickly get a full understanding of each of the units we reviewed, expediting their future planning.
