Flotek's Data Debacle
One of the big stories from last week was Flotek's data debacle that cut it's share price from $18.10 to 9.01. The company created a product called FracMax to demonstrate the economic returns of the company's complex nano-fluid (CnF).
A fund manager from Bronte Capital describes the inconsistencies in a Flotek investor presentation that provided specific operator well names and numbers based on FracMax.
That will probably be the last time you see specific well data in an investor presentation.
As an independent market research and data company, we can attest to the lack of transparency in oil and gas information. Our clients have taken rigorous steps to compare our data to public sources, internal sources, and other vendors. We take great pride in understanding the information, maintaining a high-level of quality, and monitoring the changing regulatory impacts. A difference that our clients see in our service, research, and data. Even though the FracMax production data is understated, the company is accused of providing faulty information. This story is continuing to develop, but unfortunately, this will likely become an industry issue.I would have missed this whole story, but Mike Filloon has a long article over at Seeking Alpha on this very issue. I skimmed through it; I wasn't particularly interested. Posted for the archives.
From Filloon:
Summary:
- Bronte Capital did an excellent job of identifying incorrect well data by Flotek, but may have missed the mark as to how its well performed longer term
- the Flotek well produced 24% more oil and 36% natural gas when compared to the average of the other three over an 11-month period
- when doing well comparisons, it is very important to report on a per-foot basis. Lateral lengths always differ, and this better represents well performance
On Monday, Flotek shares got crushed to the tune of 20% on a blog post by John Hempton of Bronte Capital. The stock continued lower, dropping to almost $8 from the $18 level traded last week.
This post caused quite a stir, as it directly contradicts Flotek well production data on wells using its "Green" frac fluids.
Flotek insists its technologies have shown a consistent improvement in production when compared to wells in close proximity using other fluids. We have seen this in several presentations.
Bronte has pulled the production data from Texas, and directly refutes Flotek's numbers. It has gone as far to say that Flotek's data "looks like it was made up or at least systemically rigged."
This is a big deal, as the company has other revenue streams, but the business lives and dies with patented Complex nano-Fluid® technologies. Several law firms have announced investigations. Bronte Capital is a respected firm, and although we didn't own any Flotek shares when this was announced, we decided to crunch the numbers and check the validity of its claims.Filloon concludes with two important issues. The first issue helps those trying to understand the Bakken:
When we break down production per foot, we see a much better result from the CnF well. The problem is a lack of data on well design, as we don't know the most important issue - well cost. We don't know stage length or the number of total stages per lateral. We also do not know the type or volume of proppant. Most importantly, we don't know whether the same number of frac clusters were used, or if any of these wells were high-density fracs. One could guess the Sabine wells were somewhat consistent, but this is still a guess. The Berger Unit is a Devon well, so there could be a completely different completion style. Another issue is lateral placement. The Eagle Ford interval is very thick and produces differently at differing depths. The Targac well is approximately 500 feet shallower. This could affect production, as well pressures increase with depth. These inconsistencies make it difficult to compare specific locations, but this is seen in most well comparisons. Wells production can change significantly from one section to the next, and we don't really know how good an area is until several locations are completed. That said, a conclusion can still be made with the data we have. The average three non-CnF wells produced 14.1 Bo/ft. and 26.96 Mcf/ft. over the first 11 months of well life. The Flotek CnF well produced almost 24% more oil and just under 36% more natural gas per foot. One could argue the shorter lateral length generally produces better production per foot as the operator has better control over the entire lateral length. But it still out produced the Gillespie well on a per-foot basis, and it was 172 feet shorter and used 2,042,250 gallons more of fluids.The second issue has to (more) with how Bronte interpreted things, and how Bronte may have come to those conclusions:
In summary, the Bronte blog reported a significant issue with how Flotek was reporting well performance when comparing CnF to non-CnF wells. These issues are a big deal and could be bearish for the stock in the short term. It's impossible to know whether these were mistakes or intentional, but if we focus on production, we see very good numbers from the CnF well. It is important to realize every well is different. This includes lateral length, stage length, and proppant/fluids volumes. Stimulation techniques are also important and can affect production. A location's total production is important. It is also important to understand lateral length. We cannot compare a 10,000-foot lateral to one that is half that length. Well costs increase with each additional foot drilled and completed, as does production. Breaking down production per feet helps to solve the difficulty in differing lateral lengths. This was important in showing the improvements in production with respect to the Flotek CnF well. It is also very important to be forward-looking. It is great to see a well produce large volumes of resource in a short period of time, but if a huge decline rate is seen, payback times can suffer. Although Bronte made some very good points in its assessment of Flotek and the incorrect production numbers stated, it completely missed on its assessment of Molnoskey 1H.
