Big Bad Basa

This week, we’re going to change things up a little bit.

I recently started doing research reports for current clients as well as potential clients.

They seem to be resonating fairly well so I figured we could start reviewing them together.

The intent here isn’t to make anyone look good or bad.

The intent is to explore some interesting data together, see if we can find some interesting ideas, patterns, or future analyses to conduct.

Today, we’re going to start with Basa Resources.

What are we looking at?

Before we jump in, let’s orient ourselves.

We’ll be looking at a series of graphs all fed by an underlying transaction dataset.

This dataset shows transactions between producers and purchasers along with important transaction attributes like volume, $/bbl, gravity, and more.

The point of this dataset is that it allows us to see what everyone else is getting paid for their production.

Which gives us the ability to determine if we have good or bad pricing (and remedy it if necessary).

First, we’re going to look at how Basa competes in the broader market.

Then we’ll go into a little more detail to see where there is room for improvement.

Performance Distribution

This chart gives us a way to compare Basa’s performance against good and bad producers (as measured by price delta).

Price delta means the distance your $/bbl is from the ‘perfect price’.

The closer to zero the better (more negative = more purchaser margin).

From the chart above, we see that Basa is roughly an average performer.

Nothing severe but nothing to write home about either.

Perfect price is calculated from standard pricing formulas in a table like this (you might need to zoom in):

So we can see some average performance.

But that doesn’t really help us without finding the specific areas this underperformance is caused by.

The above chart is one of 4 (I’ll spare you from looking at all of them).

The boxplots are price delta distribution the rest of the market.

The red diamonds show where Basa falls in that distribution for specific month-county combinations.

The line in each box is the median price delta - anything above it can be considered better than anything below it.

Immediately, we see some underperformance in 2024-12 in Howard county and 2024-12 Andrews county.

This presents a great opportunity to dig in and look at the individual transactions to uncover more details.

We’re not going to do that here (or else this newsletter will be 4,000 words long) but those are the natural next steps.

Show me the money

Ok so now we’ve discovered what locations and timeframes might be causing the underperformance.

A reasonable next question to ask would be “is this problem even worth solving?”

Which is fair, because Basa makes millions of dollars a year so fixing a $10,000 problem might not even register with them.

To sort this out, we’re going to look at every transaction in the same ‘cohort’ as Basa.

Cohort means producers with similar attributes like volume, gravity, location, etc…

We do this to make sure we’re comparing apples to apples.

If we look at every other transaction in the same cohort, we can isolate the transactions with higher $/bbl values and then run a sensivity analysis (where we assume that Basa got that higher price instead of their actual price).

If we roll up these numbers, we can estimate how much Basa might have to gain by improving their price.

Not bad for a month’s work.

Remember how Basa is roughly in the middle of the price delta distribution?

The 75th percentile bars show the gain at the next ‘tier’ for them.

In other words, if Basa were to improve their prices to that of 75th percentile competitors, that’s the revenue increase they might realize.

So, it looks like we have a problem worth solving since the worst one month savings is ~$50K.

Now, what do we do to solve the problem?

Data, data, profit?

The simple approach (which is always preferable to me) would be to look at the underperforming counties, isolate transactions in the same cohort with better paying purchasers, and give them a ring.

I’m well aware that purchasing agreements have different time periods, someone the size of Basa will be juggling a bunch of agreements, etc…

The point here is that if someone is getting paid more than you for the same barrel, it’s probably worth 2 phone calls:

One to your current purchaser

One to alternative purchasers

Even if your agreements aren’t up for renewal right this instance, you now have the proof and leverage available to get the right price.

Extra credit

That’s the crux of the report (I’ve condensed 27 pages of information for your sake).

However, I’m putting some other interesting data at the end in case the data-curious of you want to continue.

First, it might be good to understand what other leverage we have available for renegotiating prices.

The above charts shows us a measure of competitor density by county.

Anything above ~4000 HHI is considered very competitive.

Here’s the implication:

1. More producers == more supply

2. More supply typically means more demand (purchasers) will step in to absorb the extra supply

3. More purchasers usually means wider price spreads

4. Wider price spreads means more opportunity for us to find better pricing

We can use the counterfactual to think through this.

If we were in a county with 200 producers and one purchaser, would it be easier or harder to find better pricing?

So, counterintuitively, Basa might want to prioritize high competition counties where they have bad pricing.

Next, let’s find what counties we can bully people around in:

The idea is here is that as your proportion of a market increases, your negotiating leverage goes up.

If you are 80% of the barrels in a county and you change purchasers, your previous purchaser is going to take a serious revenue hit.

Next week we’re going to cover how moving quickly on new permit data can lead to a competitive advantage.

Talk soon,

Taylor

Industry News