Kodiak Gas Services, Inc. (KGS) Q3 2026 Earnings Report, Transcript and Summary

Greetings, and welcome to the Kodiak Gas Services Third Quarter 2026 Earnings Call. [Operator Instructions] As a reminder, this conference is being recorded. I'd now like to turn the call over to your host, Graham Sones, Senior Vice President, Corporate Finance and Treasurer. Thank you. You may begin.

Good morning, and thanks for joining us for the Kodiak Gas Services conference call and webcast to review our third quarter 2026 results. Joining me from the company today are Mickey McKee, President and Chief Executive Officer; and John Griggs, Executive Vice President and Chief Financial Officer. After my remarks, Mickey and John will share their thoughts on the current industry dynamics, walk through our quarterly results and updated 2026 outlook, and then we'll open the call for Q&A. A replay of today's call will be available by webcast and phone through November 23, 2026. Replay details are on the Investors tab of our website at kodiakgas.com. As a reminder, the information discussed today speaks only as of November 9, 2026, and may no longer be accurate by the time you listen to a replay or read the transcript. The comments made by management during this call may contain forward-looking statements within the meaning of the United States federal securities laws. These statements reflect management's current views, beliefs and assumptions based on information currently available. Although, we believe the expectations referenced in these forward-looking statements are reasonable, various risks, uncertainties and contingencies could cause the company's actual results, performance or achievements to differ materially from those expressed in the statements made by management, and management can give no assurance that such statements or expectations will prove to be correct. The comments will also include certain non-GAAP financial measures. Details and reconciliations to the most comparable GAAP measures are included in our earnings release, which can be found on our website. Now, I'd like to turn the call over to Kodiak's President and CEO, Mr. Mickey McKee. Mickey?

Thanks, Graham, and thanks, everyone, for joining us today. Before I start the call, I'd like to discuss a few changes. At Kodiak, we take great pride in developing our employees, and we are happy to provide opportunities for career advancement. Graham, who's done a tremendous job leading our Investor Relations over the last 4 years, is moving into a new role as Senior Vice President of Corporate Finance and Treasurer. Congratulations to Graham, and I have the utmost confidence that he'll be successful in this new role. Kevin Smith will now be leading our Investor Relations efforts. Kevin is a former research analyst and has over 10 years of Investor Relations experience. Many of you already know Kevin, and we believe this change will be seamless. I hope you will join me in congratulating both Kevin and Graham. Now, as we do in all meetings at Kodiak, I want to start with safety. Before each job at Kodiak, we require our technicians to perform a Job Safety Analysis or what we call a JSA. The JSA requires our technicians to review each step of the job, identify potential hazards, determine how to mitigate them and confirm that they have the right tools, protective equipment and training to undertake the work. This process requires our workforce to methodically review the scope of work and plan out how they're going to complete their tasks successfully and safely. I encourage everyone to implement their own JSA process as they are about to undertake a task, whether it is routine or something you've not done in a while, take a second to stop and make certain you understand all of the potential risks and that you have the right tools and training to be successful. Over the last 6 months, there's been an increased focus on the impact that high oil prices, diesel and even lube oil have had on the economy, creating inflationary pressure and hurting margins. Kodiak is not immune, but as John will discuss later, we've taken steps to partially offset higher fuel and lube oil prices to protect our margins and cash flow. However, what is not getting the attention that it deserves is the fact that domestic natural gas prices are down year-over-year, reducing inflationary pressures and providing the U.S. a tremendous energy cost advantage compared to the rest of the world. Given the step-up in natural gas demand from Gulf Coast LNG facilities and gas-fired power demand, the current price environment would not have been possible without investment and build-out of infrastructure assets in the Permian Basin, enabling the tremendous gas production growth. Through the first 3 quarters, Kodiak has added over 120,000 new horsepower, the vast majority of which went to the Permian to keep up with the strong production. Independent third-party estimates are reporting Permian gas growth of approximately 2 Bcf per day since this spring, and we are encouraged by various E&P industry leaders' outlooks for further natural gas production growth. Just as important, we and the natural gas industry have been able to deliver this growth in a safe and reliable manner, working hand-in-hand with regulatory agencies and the local communities in which we operate. Turning to power. It has been a volatile last few days. So I'd like to take a step back and review some of the benefits of the U.S. leading in the development of artificial intelligence. First, the advancement of this rapidly evolving technology is spurring economic growth throughout the U.S., helping drive construction, power and manufacturing jobs. At Kodiak, we've been investing in training and have been actively hiring new technicians to help develop our workforce. Last week, the U.S. Bureau of Labor Statistics announced that labor productivity increased at an annualized rate of 3.6% in the third quarter, marking the fastest pace in a year. The step-up in labor productivity was largely driven by investments in technology like AI. We are realizing these benefits at Kodiak as our technicians are using AI to improve the efficiency and safety of our operations. Lastly, we are seeing increased commitment by data centers to be good stewards in their communities by sourcing their own power and offering benefits to the local communities such as donations to build local community centers, funding for school districts and funding for trade schools. At Kodiak, we remain committed to positively impacting the communities in which we lived and work. Additionally, we focus on power assets that have minimal to no water consumption, which combined with the state-of-the-art emissions and noise abatement technology minimizes our environmental footprint. Our goal with power is the same as what the industry has achieved with natural gas compression, increasing supply to meet demand in the most safe and responsible manner. As discussed by many of the hyperscalers during this earnings season, compute capacity remains in high demand. That demand level continues to be reflected in our commercial discussions. I am extremely excited about the progress our power commercial group is making as we narrow our focus on projects that are in advanced commercial discussions that are located across multiple states. The projects that we are pursuing have the potential to rapidly put to work our new units and deliver attractive returns. In order to help prepare for the growth and technical challenges that we face today and into the future, I'm excited to say that we recently held a ribbon-cutting ceremony for our new BEARS Academy training facility in Midland, Texas. The world-class training facility provides hands-on equipment training for both compression and power assets. The 34,000 square foot facility provides our technicians with the training and knowledge needed to fully operate and maintain electrical equipment as well as perform troubleshooting and maintenance on the mechanical equipment. Through the first 9 months of the year, Kodiak employees have completed over 67,000 hours of training. This is more than just a training facility. It is an investment in the skilled workforce needed to support America's growing energy infrastructure, and we are very proud of it. In fact, we are planning an event next year to host an Investor/Analyst Day in Midland so that many of you can visit and can see Kodiak's industry-leading training and operational support facility and meet some of the people who run it. The timing of the opening of our new facility could not be better as we ramp up our technician workforce to prepare to significantly grow our power infrastructure business. We recently began site work and started setting equipment at the previously announced data center in West Texas. In total, we will deploy 40 reciprocating natural gas fuel generators with start-up expected during the first quarter. The West Texas data center provides some unique opportunities as we corral reciprocating units from 7 different existing sites on short-term contracts as well as layer in some new equipment to help meet the demand. As I previously stated, we're making a lot of progress on our power commercial efforts. Over the last several months, we've been engaged with both hyperscalers and data center developers working through detailed engineering on several projects. While speed to power remains important, we believe the industry is shifting to focus on the power providers' ability to operate the equipment and deliver a high level of reliability. This is a key selling point for Kodiak. Anyone can provide equipment, our advantage is the experience, people and ability to operate reliably every day. Lastly, John and I can't emphasize enough that the strength of our balance sheet is also a meaningful strategic advantage. Now switching to compression. Demand for contract compression remains strong as Permian gas production continues to exceed expectations. The compression market remains tight as lead times for new Caterpillar 3600 engine classes have stabilized but remain close to 190 weeks. We continue to find ways to increase our operating efficiencies through the use of AI technology, allowing us to remotely monitor our units, identify potential issues before they disrupt operations and make certain that our technicians have the right tools and training for the job. We're making great progress on signing the last remaining equipment we have available in 2027, and we have started contracting equipment into 2028. Contract compression continues to be a very high-margin business for us that generates a tremendous amount of free cash flow. This morning, we released our third quarter 2026 financial results. I'll hit a few highlights, and then I'll let John go into more detail. In the Compression Infrastructure segment, we ended the third quarter with 4.5 million revenue-generating horsepower. We drove yet another increase in fleet utilization to 98.3%. And for the first time since we closed the CSI acquisition in April 2024, our average horsepower per revenue-generating unit is back over 1,000 horsepower. Both our utilization and our average horsepower remain industry-leading metrics. In Q3, we delivered strong year-over-year growth in compression infrastructure revenue and adjusted gross margin. For the quarter, compression infrastructure revenue increased 7% year-over-year. The strong pricing performance reflects the positive progress we've made in recontracting our existing fleet and the underlying demand for contract compression in this tight market with highly visible gas growth. Compression infrastructure adjusted gross margin was 70%, marking the third consecutive quarter at or above 70%. This result would not have been possible without our unrelenting focus on large horsepower compression assets, combined with our investment in training and technology. This quarter's margin results were especially impressive considering the negative headwind we faced in the quarter from higher lube oil prices. Our supply chain team continues to do a tremendous job planning ahead and negotiating favorable contracts to help us source cost-efficient supplies and reduce price risk. In fact, we just executed a contract to lock in the vast majority of our lube oil purchases at an attractive fixed price for all of 2027, effectively taking the cost volatility off of our plate. For Power Infrastructure, we generated a 28% sequential increase in revenue as we expanded commissioning operations at the data center and short-term rental agreements were extended, staying on the site longer than expected. Adjusted gross margin for the quarter was 75%, exceeding our expectations as we were able to extend high-margin contracts and reduce our dependence on third parties. Keep in mind, in the fourth quarter, we are pulling back equipment for the West Texas data center project, which will temporarily lower some margins. Overall, our team is doing an excellent job of lowering cost and improving operations, which we expect to be highly visible in our 2027 results. We exited the quarter with a fleet of 413 megawatts. We expect to receive approximately 16 new megawatts of reciprocating power in the fourth quarter as some delivery dates for new power equipment have shifted into 2027. The market remains extremely tight. We plan to provide an update on our next earnings call about the timing and outlook for power deliveries in 2027 as we roll out our 2027 capital plan and full year guidance. Strong results from each segment drove adjusted EBITDA of $224.5 million for the quarter, up 29% year-over-year and a new company record. In summary, our large horsepower compression remains in high demand as the industry races to keep up with the rapid natural gas production growth in the Permian. Our operations team continues to do an excellent job increasing operating efficiencies, allowing us to offset the increases in lube oil and maintain high margins. The outlook for the next few years continues to be strong, and our commercial team is doing a wonderful job contracting new horsepower. On the power side, over a year ago, we sat around and completed our own form of a JSA, outlined the risks and listed the tools and strategy that we need to be successful. We've ordered the tools and equipment that we need. We have the training facility to teach the next generation of technicians to safely and consistently deliver the reliability our customers have come to expect, and we have a commercial team that is making great strides. The demand we see from AI frontier labs, hyperscalers and data center developers has only increased in the last few months as our backlog of high-quality commercial opportunities is growing by the day. These are very exciting times for Kodiak, and our outlook has never been more energized. In light of the great results in Q3, and as John will discuss, we are raising the midpoint of our full year adjusted EBITDA, compression infrastructure gross margin and discretionary cash flow guidance to reflect our increased visibility for the remainder of the year. We remain excited about the investments we are making today that will drive growth and enhance Kodiak's margins in the years to come. Now, I'll pass the call to John to further discuss our financial results and our revised outlook for 2026. John?

Thanks, Mickey. We had yet another standout quarter with particularly impressive segment margins in both compression and power infrastructure. A lot of things came together nicely to produce these results, but significant credit goes out to our operations, HR, training and IT teams for working together in harmony to continue to execute at a very high level and to continue to deliver further operational efficiencies. Let's review the quarter. We reported revenue of $402 million, up 25% year-over-year. Most of the growth is attributed to the addition of our Power Infrastructure segment, where our Compression Infrastructure segment grew nicely as well, driven by continued additions of new horsepower at leading-edge rates alongside price increases in legacy equipment. Revenue-generating horsepower increased by approximately 102,000 quarter-over-quarter, including the 43,000 horsepower we bought out of operating leases that we highlighted on our last call. We exited the quarter at $23.51 per ending horsepower, up 3% year-over-year, but down sequentially, an anomaly caused by the aforementioned operating lease buyouts, combined with the termination of a high revenue, low-margin compression contract early in the quarter. Absent these 2 events, given the constructive pricing environment for new horsepower sets and contract renewals, we expect to see that dollar per horsepower figure revert back to a steady upward trajectory as we exit the year and beyond. One of the bigger positives of the quarter was that we continue to deliver on compression segment margins. Compression infrastructure adjusted gross margin for the quarter was 70%, up 170 basis points year-over-year and flat quarter-over-quarter despite a meaningful sequential increase in lube oil costs. Our ability to produce consistently strong segment margins while experiencing a roughly 30% cost increase since Q1 on such a significant portion of our variable cost structure just simply wouldn't have been possible without high-quality execution by the teams operating those units. As McKee mentioned, by virtue of a novel supplier arrangement, we've eliminated nearly all of the month-to-month lube oil price volatility going forward. Now turning to our Power Infrastructure segment. For the quarter, we generated revenues of $42 million and an adjusted gross margin of 75%, well above the high end of our annual margin guidance. Revenues for this segment increased sequentially by 28%, driven by the expansion of commissioning work at one of our data center projects as well as the extension of a few of our high rate short-term contracts. Our power assets remain in high demand as our short-term contracts continue to be extended. At quarter end, our power fleet stood at 413 megawatts with 382 revenue-generating megawatts, resulting in utilization of 92.5%. Adjusted EBITDA for the quarter of $224.5 million was yet another company record, and net income was $71.5 million or $0.70 per diluted share. Let's turn to capital expenditures. Maintenance CapEx was approximately $21 million in Q3, in line with expectations. Other CapEx was also in line with expectations at $11 million. Compression infrastructure growth CapEx was $111 million, $34 million of which was associated with the termination of the operating leases. We had our largest quarter of the year for new fleet additions of roughly 64,000 horsepower. Power infrastructure growth CapEx declined sequentially to $68 million, largely based on the timing of our turbine payment milestones, but also due to some of our 2026 deliveries moving into 2027. Discretionary cash flow for the quarter was $158 million, up 36% year-over-year. Moving to the balance sheet. Total debt was approximately $2.8 billion at quarter end. Our primary equity offering in May generated approximately $836 million after expenses, allowing us to prefund the equity portion of our growth capital spending plans. Our leverage ratio at quarter end was 3.1x, and we had approximately $1.6 billion of unused revolver capacity. Going forward, we think that the strength of our balance sheet plus our access to multiple sources of relatively low-cost debt capital provides us with a meaningful competitive advantage as we build out the power leg of our energy infrastructure platform. Finally, our Board declared a dividend of $0.51 per share, a 4% year-over-year increase that will be paid later this month. Based on our third quarter discretionary cash flow, coverage exceeds 3x. I'll wrap things up with some commentary on our updated 2026 guidance. Though we made several adjustments, I'm only highlighting the more meaningful ones. First, we raised the low end of our compression infrastructure adjusted gross margin, resulting in a new range of 69.5% to 70.5%. Given the heavily contracted take-or-pay profile of our compression business, combined with good visibility in our largest cost items, we're very comfortable with our guidance at this point. Within the Power segment, we elected to move up the bottom end of the adjusted gross margin range. Outperformance in Q3 is expected to more than offset the impact of repositioning some of our working fleet to prepare for the Q1 West Texas data center project. As a result of these updates, we raised the midpoint of our overall adjusted EBITDA guidance to $850 million. In terms of growth CapEx, we increased the midpoint of our compression infrastructure growth CapEx by $10 million. We've been working with packagers to accelerate new unit deliveries where we can to help us better align with customer development plans and overall demand. We think we'll add around 170,000 horsepower for the year, including the roughly 21,000 horsepower from the Permian purchase leaseback we closed in the first quarter. For power infrastructure, we're reducing our growth CapEx forecast to $350 million to $400 million, which reflects some new power unit deliveries moving from 2026 to 2027 as well as a more firmed up payment schedule for balance of plant items. We now expect to receive approximately 16 megawatts of new recip generators in the fourth quarter for a total of 29 megawatts for the year. With that, I'll hand it back to Mickey.

Thanks, John. I'd like to conclude by coming back to the concept of the job safety analysis that I discussed at the beginning of the call. Successful outcomes don't happen by accident. They require preparation, the right equipment, the right people and a disciplined approach to execution. At Kodiak, we believe we have those ingredients in place. Our compression business continues to benefit from strong market fundamentals and increasing demand for large horsepower equipment. At the same time, we are investing in the assets, training, technology and talent necessary to build a leading power infrastructure platform capable of meeting the growing needs of our customers. The opportunities in front of us continue to expand, and our confidence in Kodiak's long-term outlook has never been stronger. Most importantly, we remain focused on executing safely, operating reliably and creating long-term value for our customers and shareholders. Thank you for your participation in today's call and your continued interest in Kodiak. With that, we're happy to open up the line for questions. Operator?

[Operator Instructions] Our first question comes from the line of Elias Jossen with JPMorgan.

I wanted to start on the power infrastructure outlook. You highlighted positive conversations with counterparties in your remarks, but just want to get a sense if we've seen a shift following the political developments this past week. Can you just kind of address the tone of those conversations and if they've changed post election and maybe more specifically, if there's kind of a shift in time line or willingness to continue development in Texas overall?

Elias, this is Mickey. We're having lots of those conversations going on right now. It is early days following the election. But the overall tone of the conversations we've been having is that there's not much change in development plans. I think that there is some -- a little bit of uncertainty in Texas, but Texas isn't the only place that we're looking for -- looking at these opportunities in. And -- I think that as we move forward, as I said in my prepared remarks, that the demand for compute capacity is still extraordinary right now. And I think that the development of these will continue to happen and the build-out will continue. And whether it's in Texas or somewhere else, we're pretty agnostic to it. So we can take this power wherever we need to.

Awesome. I appreciate the color. And then maybe just taking that a step further. You have a really strong backlog of equipment. It seems like that's increasingly valuable in the compute-constrained world that you frame. But just on the timeline and the deployment, have we seen a shift? Do we kind of still expect those previous targets you laid out? Or how might that evolve?

We think we're still pretty firm on our deployment ability here. The bulk of our turbines that we've ordered from Baker Hughes don't start coming in until the fourth quarter of 2027. So, I think a lot of the questions will be answered well in advance of that, and we'll have the ability to make sure that we can get those deployed at attractive prices and terms.

Our next question comes from the line of John McKay with Goldman Sachs.

I want to start on the kind of core compression side. A year ago, you guys at least gave kind of soft guidance for 8% EBITDA growth for compression through 2030. I think over the last year, we've seen Permian gas expectations go up. We've seen lead times for compression get longer. You've done a really good job on the cost side. Any puts and takes on that 8% when you're looking at that a year later now?

Yes. John, it's John. I'll take that one. So look, a while back, we gave our 8% or our growth algorithm being that 8% compounded for many years from an EBITDA perspective from the compression business. Then we went further, as everybody knows, and we put out our kind of 150,000 horsepower per year for the next 5 years to wind up at $5.2 million. When you factor in some of these purchase leasebacks and operating lease buybacks, that $5.2 million is now mathematically about $5.3 million. So here's our answer. We couldn't be more excited about the compression business going forward. And whether it's the E&P earnings releases and there talks about gas production in the Permian or if it's the FIDs that have been green lit over the last quarter, the Permian is going to produce a massive amount of gas going forward, and it's going to need a massive amount of compression. So I guess, if pressed, I'd say be over on the $5.3 million. It feels like it's going to be there at great returns going forward, and we just hope we can take care of our customers.

Yes. I'll just take that a step further a little bit there, John, and just say we've got more certainty around lube oil pricing going forward. We know our growth going forward. We've got the bulk of our fleet locked up under long-term contracts. We understand the escalators that are in those contracts and our ability to redeploy equipment. We feel really good about the outlook there.

I appreciate that, guys thank you. And maybe just staying in compression, not to get too far in the weeds, but pricing in the quarter, John, you touched about -- touched on a couple of the moving pieces. But do you mind unpacking it just a little bit more for us in terms of maybe ex some of the noise, what growth could have been and then not to get ahead of '27 guidance too much, but just generally speaking, on how we can think about pricing going into, let's say, the next year?

Sure. I'll take the first piece and hand it back to Mickey. So we printed $23.51 per horsepower, which was a step down slightly from the prior quarter. We previewed with anyone that was listening before that we would expect a step down. And I called out 2 things in the prepared remarks. Number one was when we bought out operating leases, which basically increased the denominator and dropped the dollar per horsepower. And then secondarily, early in the quarter, we had the termination of really one of our lowest margin compression projects out there, too. And so if you eliminated those 2, we have been just a shade over $24 per horsepower. And I guess the message that I leave for you and any investors is that anyone that kind of reads through this and says that there's pricing weakness will be disappointed when we report next quarter to get back on that upward trajectory. And then Mickey, if the second question was about outlook on pricing going forward.

Yes. And I don't think we've seen any -- any change on outlook on pricing going forward. If you look at the contracts that we actually put in place in Q3, you saw -- we saw a sequential step-up in that pricing on those. We don't report that publicly. But you did see a lot of horsepower get set in Q3, much of which only provided a partial quarter's revenue contribution. So you'll see the full run rate of those in Q4 also. And so we still feel really good about kind of the numbers that we've always talked about here in the compression pricing and no change there.

Our next question comes from the line of Praneeth Satish with Wells Fargo.

First, congrats to Kevin and Graham on the new roles. Maybe just on the power side, looking at the projects in the funnel that you have out there, can you comment on the contract duration that you're seeing? What's the minimum tenor of projects that you're seeing out there and the range? And then also, I know, Mickey, you mentioned that there's projects outside of Texas as well. Is there any way to kind of break down for the funnel what the geographic mix is and what percentage is perhaps in Texas versus outside of Texas?

Yes, sure. Thanks for joining us, Praneeth. I would say just to start on the second part of the question, probably the majority of the projects we're looking at are in Texas, but we are looking at multiple states around the lower 48 here. We feel pretty good about the ones in Texas continuing to move forward here. I think you've seen some softening in some language. And as we talked about in the prepared remarks, right, water usage and noise abatement and emissions controls are going to be really key to those being built out in Texas, and we feel like we've got a good handle on those things. As far as contract term, we're not really ready to announce one yet publicly, but we're seeing a minimum of 10-year terms, most of them at 15-year terms and there's no change to the contract tenors there that we're seeing in some of those commercial discussions.

Got you. That's helpful. And then I know you're not providing 2027 guidance yet, but maybe just conceptually, when we think about the power business in Q1, I guess, is the major moving variable here, the timing of the West Texas data center commissioning. Is there any other kind of puts and takes we should think about? And then once that facility is fully commissioned, would the power revenue look kind of like the Q3 revenue run rate? Is that the right way to think about it?

I think pretty close. That's probably right. I mean, like I said, we're pulling back equipment right now to be able to deploy to that data center in West Texas. And so you'll see a little bit of a dip in Q4. But I think we'll be right in that ballpark of that kind of Q3 run rate when we get that stuff up and running.

Our next question comes from the line of Doug Irwin with Citi.

Maybe a bit of a follow-up on the power margin and contracts there. You've already seen some success improving the margin on the existing business. Could you maybe just talk a bit about where market rates might be relative to some of the legacy contracts you inherited and where you see margin trending from here once you layer on more contracts? And then is there still more equipment in the field that needs to be termed out? Or should we expect most contracts from here to be tied to new equipment?

Yes. Thanks, Doug. I appreciate you joining us today. I think that as far as legacy contracts, they're a little bit all over the board, to be honest with you, that we inherited. Some are underpriced. Some of them are short terms that are kind of overpriced, to be honest with you. just because those were on like 3- and 6-month type contracts. And as GPS was looking to maximize their cash flows in the short term for sure. But I think that, that margin, you'll see a dip again in the fourth quarter, like I just said, because of the -- we don't want to -- we may pull back equipment and not be generating as much revenue on it, but we're certainly not going to let go of the people associated with that. So we'll maintain some of that cost. But as we put that stuff back to work here in the first quarter, you'll see that recover a bit. And so -- and then I think that going forward, yes, there is quite a bit -- I wouldn't say quite a bit, probably a couple of hundred megawatts that we have to put -- pull off of legacy contracts and put back to work and looking for the right places to do that around the country right now.

Got it. That's helpful. And maybe a follow-up just on the opportunity set beyond data centers for power. Do these opportunities maybe become a bit more attractive here given the events of the last few weeks? And if so, how should we be thinking about the potential scale of these types of opportunities relative to some of the data center deals you might be working on?

Yes. I mean we're always seeing additional opportunities in the microgrid world, especially around the Permian Basin for oil and gas. And we think those are really attractive. And we really haven't changed or shifted the focus at all on those. We're working on a few of those. And as long as they provide the term and returns that we're looking for and comfortable with, then we'll continue to work on those as well. So there's multiple of those in the pipeline and at attractive kind of returns and keep working hard on those and get some of those to fruition.

Our next question comes from the line of Neal Dingmann with William Blair.

My first question is just on the power contract side. Specifically, John or Mickey, just that 76 megawatt of power that you announced, I think that was in September. Has that been deployed? And then remind me how -- it seems like you've had -- continue to have very favorable pricing around this. Maybe just talk about the structures there.

Yes. So the 76 megawatts, that's the West Texas data center that we've referenced here, Neal. That is equipment that we are beginning site construction and site prep on right now and will be deployed kind of starting at the end of December and into January and February with the goal to be energized before the end of the first quarter here. I think as far as the other contracts go, I think everything that we're seeing right now that we're working hard to progress is penciling in exactly kind of where we expected it to be.

Thanks Mickey, and then second question, just a very easy one on the compression side. Is there still opportunities that you potentially see out there for the purchase leaseback from potentially some of your customers?

Yes, we are seeing some significant interest there and opportunities that we're working on. And those are a little bit hit and miss. That has a lot to do with kind of corporate philosophy and cash flows of our customers. And so we're looking at a few of those and hopefully continue to progress on those conversations. But that opportunity is definitely still exists out there.

Neal, this is John. I wanted to go back to that. Yes, I just wanted to quantify to that pricing just because I want to make sure it's out there on a project level basis in the power business, all of our projects are penciling in at high teens or better with, call it, somewhere between a 5- and 6-year payback cash-on-cash return. And that's been consistent since we started these conversations, and it seems to be what we're seeing on our key negotiations.

Our next question comes from the line of Derek Podhaizer with Piper Sandler.

Maybe just to go off on some of those last comments. Just want to get some insight into your most recent conversations with your hyperscaler and data center developer customers on maybe the potential scope of the projects increasing or even pulling in some additional services that we hear about from some of your peers and maybe this drives up the dollar per megawatt of capital cost. I think you guys have mentioned around the 1.5, but I would sure pricing would go along with it in those economics that you just laid out, John. But any insight as to far as are these projects getting bigger? Or is the scope expanding and what that can potentially do to some of the upfront capital that you have to deploy?

Yes, I think that we expected from the get-go to be fully responsible for balance of plant and potentially batteries -- battery backup for some of these things. So I wouldn't say that the scope is changing much from our perspective, but it is something that we were expecting going into this, and it's going to be something that adds to the total capital of these projects. So you are seeing a little bit of difference in the market and some of these data center developers that we're talking to actually want to own the batteries, but outsource the power and balance the plant. So it's a little bit of difference depending on who you talk to, but we absolutely are going to do exactly what you'd expect and what the others are saying out there and provide a full turnkey solution here from construction to start-up to operations.

Great. That all makes sense. And then just on compression, a lot of conversation on pricing as far as the trajectory there. But can you touch on the term that you're seeing? Any changes in some of the leading edge term? Are you seeing that extend out just given the tightness in the market? I think you said CAT is still at 190 weeks or so for deliveries, and I think potential air pocket hitting us in the back half of next year. Are your customers willing to sign up for more extended term? Any thoughts around that?

You're certainly seeing customers that are willing to talk about more extended terms today in that 5, 7 and even 10-year range. So it's not super consistent, but definitely, the tenor of the conversations definitely is longer than it was just 2 or 3 years ago.

Our next question comes from the line of Theresa Chen with Barclays.

I'd also like to congratulate Graham and Kevin on your new roles. John, I want to go back to your comments about compression segment margins. How much of the compression cost structure remains exposed to commodity-driven fluctuations at this point? And as you mitigate this exposure through longer-term procurement, how do you think about the trade-off between securing greater cost visibility versus maintaining flexibility if market conditions change?

Yes. Theresa, as of today, none. I mean it's like -- it's very, very minimal by virtue of what Mickey mentioned in his call before, too, and that we've entered into an agreement with our supplier to lock in our lube oil prices for really all of '27 and starting with December. We already had a contract, which gave us visibility through most of this quarter anyway. We -- the compression business is very stable given that it's production oriented. And so our volumes are really well known. And so the only variable that we really have to control against is price. And so now we've taken that basically off the table. So we love it. We've got enough issues to kind of deal with business day-to-day as opposed to having to deal with that commodity price fluctuation. So we think it's going to allow us to at least lock in a budget and continue to work on capturing these training and technological efficiencies, and that's what's going to drive our margin going forward.

Understood. And going back to the Baker agreement. So if the commercialization timeline extends, how much flexibility do you have around the timing and pace of associated power capital spend? And maybe more pointedly, can you help us understand the level of capital that you've committed under this agreement? How should we think about the potential cap balance sheet exposure over time? And does the recently announced Anthropic OnePWR deal inform your thinking about the economics of alternate options for this power supply?

Yes I bet Mickey is going to tag team this. So we're committed to our core strategy of putting this power to work with customers under long-term contracts with great returns. And irrespective of what happened last week, we're really not deviating from that strategy, and we don't see any reason to. So that said, in terms of the flexibility and like how much money we committed in the balance sheet strength and all that stuff. So we've kind of put in that press release and I think I've told everybody that we've committed to go buy, call it, 1.1 gigawatts from Baker Hughes over the next -- through 2031, I guess it is, too. That's something north of about $1.4 billion in capital with Baker Hughes. We do have flexibility like if worse were to come to worse, like we could step away from that. It would leave a mark, truthfully, but we could if we needed to. And when you reference OnePWR, for those who don't know that, Anthropic bought an independent power companies just raw power, a bunch of CAT engines at the end of the day for a nice premium relative to what they paid for it. So we do think sitting here today that power is still a scarce resource. We fully subscribe to the idea that compute is in tremendous demand and irrespective of what happened last week, like it's only going to go up. The question is where are we going to place the data centers. So I think that power is still going to be in great demand. Mickey said it, I said it in our call, we have an awesome balance sheet. The fact that we prefunded our equity needs last quarter when things were really, really rosy is great. We feel really good about riding out the next couple of years through this investment cycle that we're in and kind of living within our leverage targets that we set before, which is at 3x to 4x. And we sleep well at night as it relates to our balance sheet and our available liquidity and our ability to borrow more debt capital at attractive rates through the litany of options that we have.

Our next question comes from the line of Elvira Scotto with RBC Capital Markets.

So you mentioned it in response to an earlier question, but you've attributed a meaningful margin improvement to, I guess, AI, machine learning, technology investments that you've made. Can you maybe put a finer point on the ROI from these technology investments? And how much have these investments contributed to margin over the past year? And what are the incremental gains that you think remain to be captured as you roll out large language model tools to field technicians?

Yes. Thanks, Elvira. It sounds like I'm talking to my Board right now as well, too, because we get this question all the time. So -- and it's the same question we give our Board, too is it's not one thing, it's everything, right? And we're actually spending -- if you thought about -- I'm not going to give like raw dollars, but if we thought about the actual capital that we're spending on machine learning and artificial intelligence, it's quite low relative to, I'll call it, the overall size of the business and the incremental dollars that we're generating. But that in and of itself is only like some of the puzzle. The other piece is all the investment we've made, I'll call it in training, but it's really around people development, attracting, developing, training and retaining like these key folks and then teaching them how to deploy these great technologies. So it's not as simple as saying spend $1 here, get $3 back there, but the returns are the margins are kind of proof in the pudding, so to speak. We do think that, like I've said it before, pre-CSI in 2024 and really about the time we really hit the afterburner in machine learning, AI and all the training stuff that we're doing, we were at 66% gross margin business in our compression business and CSI was at 55%. We merged the two, we captured some synergies. But today, we're punching 70%, and that's after about probably 1 to 1.5 percentage points of overhang from higher lube oil prices, so call it 71% or so. So I come back and look at a $1.2-ish billion revenue business and think about the margin enhancement that we're seeing and the payback is really, really significant relative to the total amount of dollars that we spent. In terms of where it can go, we never thought we'd be a 70% gross margin business with our profile followed a couple of years ago. We're not going to say where it's going to go, but it feels like we've got some room to improve kind of going forward. But definitely will be more difficult every quarter we go forward because we're already at great levels.

Great. And then I know that the Permian is driving a lot of the growth in the compression, especially as gas production seems to outpace expectations there. But are you seeing growth in other basins as well?

Yes, Elvira, this is Mickey. Absolutely, we are. We just signed some nice contracts in '27 and '28 for growth in the Northeast and definitely seeing some opportunities in the Rockies as well as in the Eagle Ford and down in South Texas as well.

Our next question comes from the line of Connor Jensen with Raymond James.

Most of my questions have been answered already, but just had one quick one on the other services segment. Just wondering kind of what we should think about that segment relative to the pretty strong revenue year you've had here in 2026? And is there a meaningful amount of power-related projects that could impact that? Or how should we think about that segment going forward?

Yes, great question. We don't really talk about the segment a bunch given its relative size. But I'd say that segment is dominated by 2 pieces of business. It's our projects business, which is kind of us being a project manager and general contractor on oilfield or historically the stations or oilfield production equipment that carry relatively low margins. And then it's our aftermarket services business where we do kind of mechanical work on our customers' equipment that they own or we sell parts to them. And they're both about the same size or so. I'll start with the latter one first. On the AMS business, admittedly, we've probably been so focused on integrating CSI, focusing on compression, focusing on power that we'd underinvested in it. Now we've kind of got more emphasis placed on that business. We brought in new leadership. We're seeing some kind of great I'll call it, green shoots in terms of where that business can go going forward. And it's never going to be huge by any stretch, but it definitely can grow from where it is today. And then on -- so that looks like good to us over the next couple of years. And we think with all the compression equipment that's out there that customers will increasingly turn to Kodiak to help take care of their equipment. Pivoting to the project management business. So we did have one pretty large reasonably low-margin project that we took care of for a customer, a very strategic customer of ours during 2026 that's at the tail end of the project. And as we look forward on the, I'll call it, the non-power side, it's kind of just project timing as to when the new projects kind of will kick off. But I do want to pivot to power as we embark on kind of this capital -- the growth in the power business we do expect to really see a ramp-up in our station construction or our project management business around power. It will be lumpy because it will be associated with the build-out of the island and power plants, but it will be substantial as well, too. So I wouldn't read too deeply into kind of the Q4 piece because it's going to get back on its horse and kind of continue in '27 as the power kicks in, it's going to kick in, in project management quite a bit.

Our next question comes from the line of Julien Dumoulin-Smith with Jefferies.

This is Patrick on for Julien. Congratulations to Graham and Kevin again. I just had one left. I was wondering, I think you've spoken before about the detailed kind of engineering studies that you've been doing for the upcoming power contracts. And you also mentioned that the majority of these that you're negotiating were going to be in Texas. I'm wondering if you pivot to other states or even about a pivot to other states, if there's any change to your expected time line for deploying those into revenue-generating work if you have to restart the engineering from scratch with new sites?

I wouldn't suspect so. I think that the engineering that we've been doing is pretty applicable to other states as well. So it really is just electrical design. So I don't think that, that would -- the engineering piece would delay us too much. But we're looking at -- like I said, we're looking at projects in multiple states right now. So I feel like we're in a good spot.

Okay. And as a follow-up, I think just to clarify some of your comments, you're still very confident in Texas for the majority of those projects to continue. So you expect Texas to be where a majority of them end up or that's where your starting majority was and it might change?

Well, I think we're still in the early days following this election and the midterms here. And so there's a lot of water to go under the bridge. So we'll see what happens. But we still feel pretty confident with it that a lot of the potential customers we're talking to still feel like they're moving forward with their projected plans. So I still feel pretty good.

Our next question comes from the line of Stephen Gengaro with Stifel.

Two quick ones for me. One, just -- I'm not sure how much you'll comment, but on the power timing growth, outside of the 77-megawatt deal that goes to work early next year, should we be thinking the next kind of big bump up in the power business is early '28 when the larger assets from Baker start delivering?

Yes, I'll take that. Yes. So by virtue of when we get the turbines from Baker Hughes and then by when we could have them energized, it really is kind of the early part of '28 before you'll see the significant ramp. We do have incremental reset power coming to us during '27, which we would expect to put to work as well, too. But the significant ramp is going to be in '28.

Great. And then just the second one quickly. On the compression horsepower side, is there a meaningful uptick in the price you're paying for assets given the tightness in the supply chain? And I assume that's baked into the economics of deals, but can you comment on that?

Yes, it's absolutely baked into the economics of the deals. We have seen some inflationary uptick there, but I wouldn't say it's significant. It's in that kind of mid-single-digit range.

Our next question comes from the line of Nate Pendleton with Texas Capital.

Congrats on the quarter. I wanted to touch on capital allocation here with the recent increase in the dividend. How are you weighing return of capital versus the opportunity to invest further in both businesses or reduce leverage?

Yes, it's a great question. We expected it would come as well, too. So I guess I'd state just specifically, a $0.02 increase equals a $0.02 increase is a 4% dividend increase. And from a shareholder returns and an overall capital allocation perspective, we and our Board, we rigorously evaluate the appropriateness of that dividend each quarter to ensure we have a smart capital allocation and shareholder return strategy that works for our investors and is also sustainable over the long term. So we've got a big investment cycle that's in front of us in the power pieces everybody knows about. And we are an infrastructure company in the midst of an investment cycle. And our compression business is durable contracted cash flow. It's going to be here for the long term. And then once we get the power that we've ordered under contract, it's going to have the same characteristics only under longer-term contracts. So we think that our investors going into this were infrastructure investors that value the dividend. We think we still have many of those same investors who might have attracted some new ones. But as we come out of this, we are going to be an infrastructure company. And so we think it's important to continue to have a dividend that's kind of sustainable but growing each year. And then we absolutely think it's paramount at all times and even more so at times like this, to protect our balance sheet. So I come back to kind of where we were before. We did raise equity last quarter when it was the right time to do so in effect, prefunding our equity needs. We've got $1.6 billion of liquidity under a great revolver, which we have the ability to upsize, and we've tapped the capital markets quite successfully going forward. I think having the compression business, call it, whatever, 80% of our business today and that high-quality durable leverageable cash flow gives us a significant competitive advantage relative to others that don't have a core business that's quite as good to allow us to continue to fund our growth using debt capital at what we would say is a variety of options.

Got it. And then maybe a follow-up on just the mention of the balance sheet there. In the past, you've talked about potentially being a consolidator in the distributed power sector. With some of the market volatility that we've talked about on this call, and that's in the broader power sector, can you talk about how the opportunity set is evolving for adding assets or capabilities in that business?

Again, following these midterms here, it's a little bit unknown right now. And so we'll continue to monitor the market and see if there's opportunities to consolidate and always interested in looking at that, and we'll continue to evaluate any of those opportunities.

Our next question comes from the line of James Larkin with Bank of America.

I guess, quickly, just pivoting back to power pricing in the quarter. Revenue per megawatt was up 20%. I think we talked about it a little bit, but on -- you had some -- in your prepared remarks, you had some comments about new data center commissioning and longer time on site for some existing units. So I was wondering if you could just give a little bit more color on those, maybe what you're seeing with pricing there and longer contract durations.

Yes. I think what you see, right, is a lot of these people had power on short-term contracts. And as they think that the grid will come in to supplement their power and that kind of thing, it's slow to get there. So a lot of people aren't willing to give up the power that they've got under contract or on location right now. So lot of them trying to renew, and we're obviously weighing renewals for short-term contracts with pricing and bumping that a little bit as we go through that. So we're maximizing the value of that stuff while it's on location. So it's kind of what you'd expect that stuff is just staying on location for longer than what the contracts have it set to be. And we're continuing to manage that fleet appropriately and figure out when we can get that stuff off and get it put to work in long-term contracts.

Great. And just a quick follow-up. I just wanted to confirm, and we talked a little bit about midterm elections and post those. And just wanted to confirm that the West Texas data center site is still on the same timeline and there's no potential for anything shifting to the right there.

No, same time frame for that.

Our next question comes from the line of Zach Van Everen with TPH Research.

Maybe just on power. I know you guys have touched on it before, but maybe as conversations have progressed, are you getting a better idea of how big each of these projects will be or just the general size that you guys are seeing out there for requests?

Yes. I mean, look, I mean, all of these projects have multiple phases to them. So it's start out with anywhere from a couple of hundred megawatts to 500 megawatts and grow to over 1 gigawatt at some point in time. And so -- all of them are different, but there certainly are multiple hundred megawatts that we're in conversations with at any given time for each project.

Got it. That makes sense. And then maybe one more on the compression timelines. You guys mentioned 190 weeks. I was curious how much of this timeline getting pulled further and further out is related to the power build-out and just demand for the CAT facilities versus traditional compression and pipelines?

You know, Zach, it's both, right? I mean compression is in high demand right now, and there's a lot of -- of growth that we're looking at from gas volume growth around the country right now, especially in the Permian. But the demand for power is also straining kind of upstream in supply chain lines there, too. So it's -- the effect is from the flow from both sides there.

Ladies and gentlemen, that concludes our question-and-answer session. I'll turn the floor back to Mr. McKee for final comments.

Thanks, operator, and thanks to everyone participating in today's call. We look forward to speaking with you again after we report our results for the fourth quarter. Bye-bye.

Thank you. This concludes today's conference. You may disconnect your lines at this time. Thank you for your participation.