Prepared remarks
Ladies and gentlemen, thank you for standing by. Welcome to Talen Energy Corporation Third Quarter 2025 Earnings Call. Operator instructions were provided. Please be advised that today's conference is being recorded. I would like now to turn the conference over to Sergio Castro, Vice President and Treasurer. Please go ahead.
Thank you, Michelle. Welcome to Talen Energy's Third Quarter 2025 Conference Call. Speaking today are Chief Executive Officer, Mac McFarland; and Chief Financial Officer, Terry Nutt. They are joined by other Talen senior executives to address questions during the second part of today's call as necessary. We issued our earnings release this afternoon, along with the presentation, all of which can be found in the Investor Relations section of our website, talenenergy.com. Today, we are making some forward-looking statements based on current expectations and assumptions. Actual results could differ due to risk factors and other considerations described in our financial disclosures and other SEC filings. Today's discussion also includes references to certain non-GAAP financial measures. We have provided information reconciling our non-GAAP measures to the most directly comparable GAAP measures in our earnings release and the appendix of our presentation. With that, I will now turn the call over to Mac.
Great. Thanks, Sergio, and welcome, everyone, to today's call. We appreciate your ongoing interest in Talen. Before we review the quarter, I'd like to start with where we are as Talen and talk about the broader landscape as we implement the Talen flywheel. Sitting here today, the overall market continues in the same trajectory as we've discussed in prior calls. AI and data center capital budgets continue to impress and expand. It seems like every day, there is another announcement, investment or idea floated on how to power growing data center demand. Demand for power keeps coming, and we will need an all-of-the-above approach to solve the growth from the supply side. Pennsylvania continues to drive economic growth through data center development and remains a pro-business place to invest. Governor Shapiro, the Pennsylvania PUC and the local communities have embraced these investments and recognize the advantages Pennsylvania brings to developers, including speed to market, while at the same time, they recognize the need to properly allocate cost, build new generation and keep residential rates in check. We are doing our part by evaluating options to solve short-term capacity questions and longer-term resource adequacy. It's why we recently signed an MOU with Eos Energy to partner on battery development in Pennsylvania and PJM using Pennsylvania manufactured batteries. As I've said in the past, for the next five years, we are going to need to solve a capacity issue, not necessarily an energy issue, and we believe batteries and peaking plants can solve this need much more readily than CCGTs and at overall lower cost. CCGTs will be needed, too, and we will look to build them in collaboration with the right partners and customers, but that is further off on the horizon. And while I have focused on Pennsylvania in these comments, we continue to be excited about the prospects in Ohio, given load there already exists, but more on that in the coming quarters. At Talen, we have a number of things in flight. First, we continue to execute under our existing agreements with AWS and the Susquehanna site continues to be built at an amazing speed and has been electrified. Second, we are working diligently to close the Freedom and Guernsey acquisitions, which will add to our baseload fleet and to our large load contracting strategy. As you know, we refiled our HSR application at the Department of Justice for a second time, restarting the 30-day time line, which now runs through November 17. We believe it was prudent to give the DOJ additional time because the transaction fails zero market power screens. We remain confident that we will close these acquisitions. I do note that this might take a little longer and bleed into Q1 of 2026. That said, we are pressing to get these deals closed as soon as possible and might be able to get them done as early as late this year. We are optimistic that FERC can then act on our 203 filings in short order. In addition to pushing our regulatory approvals or pushing them along, we also recently closed on a highly successful financing package. And lastly, with respect to the acquisitions, Dale and the fossil team are well underway on the planning to add both Freedom and Guernsey to the fleet, and Chris and the commercial team are ready to fold them into the portfolio when they can. We are truly excited to get going with these assets and want to thank the Caithness team on the professional transition to date. Third, we continue to pursue additions to the flywheel through large load contracts and additional megawatts for the portfolio. On the contracting side, there has been a fair amount of noise in the markets about our ability to contract, when we might be able to contract, how we might do so and how we will manage the so-called gas risk. Trust me, when we hear all of that, we just go about our business. We built a good comprehensive playbook with the Amazon Susquehanna contract and our focus currently on execution. Our strategy remains the same. Our efforts have only been redoubled and our focus has sharpened and our commercial learnings continue to expand. We have been working on the next thing since we signed the first AWS contract in early 2024 and gained a lot of commercial knowledge by changing that contract to front of the meter. Refiling of our HSR on Freedom and Guernsey does not change our path. Let me be more explicit as to why the exact closing of these deals doesn't impact our near-term strategy. In the rest of the portfolio, excluding Freedom and Guernsey, we have approximately 4 gigawatts of gas-fired generation between Montour, Lower Mount Bethel and Martins Creek as well as 300 megawatts of carbon-free power at Susquehanna remaining. Our efforts at Montour continue, and I'm sure you have seen that we are working on zoning and permitting at the site, which, by the way, if you were with us back in 2023, was the same activity we were undertaking at Susquehanna then. All of this activity goes on, all well before we close the acquisitions. And as I've said, we remain confident that we will still close the acquisitions in short order. So timing of closing of Freedom and Guernsey is really irrelevant to our contracting power strategy. We feel good about our ability to look at further expansion of the portfolio through acquisitions and continue to explore free cash flow accretive deals. And we are constantly challenging ourselves to reshape the portfolio. And I'm sure someone will ask when we might expect to announce the next part of the flywheel. And I'll say the same thing that we always say, which frankly, is next to nothing. You'll be the first to know when we're done. We want the right deals, not any deals and not on anyone else's time line. In the meantime, 2026 is setting up well. We are reaffirming '26 guidance, and we are starting to see forwards tick up. Gas is up, sparks are expanding and load continues to be strong. All factors that continue to impact commercial positioning on long-term transactions. Terry and Chris will walk you through what we are seeing in the markets in just a bit. Moving to the present and Q3 on Slide 2. As we said during our Investor Day call, we saw limited volatility and limited opportunity to capture incremental value in the third quarter, and the quarter was a little light of our internal expectations, but we already knew that and acknowledged that in September at our Investor Day. Going into the summer, we were looking to regain some of the lost opportunity from the spring outage at Susquehanna, but it just didn't materialize. Both July and August experienced fewer peak load days when compared to June, and we experienced incremental forced outages as our fleet continues to run with higher capacity factors and longer run times between maintenance. Terry will provide a few more details on this later. During the quarter, we delivered $363 million of adjusted EBITDA and $223 million of adjusted free cash flow. So far, in Q4, things are a bit better given the market move up, but we are still projecting to be at the lower end of our guidance range as we previously stated at that September Investor Day. Before I turn this over, I'd like to thank our entire Talen team for their hard work and dedication, and I'm happy to let you know that we have reached a five-year extension of our Local 1600 contract with the IBEW. So a special thanks to Rusty and the IBEW leadership team. At Talen, we are powering the future together. With that, Terry?
Thank you, Mac, and good afternoon, everyone. Turning now to Slide 3. Mac covered the regulatory process, but let me provide an update on the financing of Freedom and Guernsey. Last month, we successfully executed several financing transactions at attractive rates to fully fund the acquisitions, including $2.7 billion of senior unsecured notes and a $1.2 billion senior secured term loan that contains a delayed draw feature. The pricing we received exceeded our initial expectations as the credit market continues to demonstrate demand for Talen paper. We also received commitments from our bank group to increase our existing revolving credit facility to $900 million and to increase our existing letter of credit facility to $1.1 billion, while also extending the LC maturity date by one year to December 2027, all to further support the impact of the acquisitions on the financial operations of the business. Congratulations to the treasury and legal teams for a successful set of transactions. We are currently earning interest on the senior unsecured note proceeds and will not draw on the term loan until the closing of either the Freedom or Guernsey acquisition. Turning to Slide 4. As Mac said earlier, nothing has changed thematically. The underlying fundamentals for Talen's value proposition remains strong. Macro fundamentals and AI demand remain intact. Last week, we continued to observe the trend of hyperscalers seeing tremendous growth in their cloud and AI businesses, which in turn has led them to continuing to raise or affirm their capital investment plans. As you can see in the chart on the upper left, the projected growth of spend from hyperscalers continues in earnest with total CapEx projected to be $700 billion in 2027. Not only is the capital commitment growing, but the acceleration of demand for power continues. For example, Amazon noted on their earnings call last week that they have accelerated capacity additions over the past 12 months by adding 3.8 gigawatts and expect to add over another 1 gigawatt in the upcoming fourth quarter. Further, they expect to double their overall capacity by 2027, which would add in excess of 10 gigawatts of capacity in North America alone. We see significant load growth coming over the next decade from hyperscalers as well as reshoring of manufacturing and the ongoing electrification of the economy. But what about the current load conditions? Q3 2025 provides a clear example of the change in the load growth in the PJM market. Overall, Q3 weather was flat compared to the same period in 2024 as measured by cooling degree days. However, the average electricity demand was higher. During the quarter, we saw approximately 3.4% of incremental power deliveries on a weather-adjusted basis in PJM when compared to the same period in 2024, a clear sign of demand growth in the market that is expected to continue. Furthermore, for the first time in over a decade, we experienced two of the top peak demand days at PJM during the heat event in June. These two demand peaks registered in at the third and fourth highest summer peak demand readings in the history of the market, further evidence of demand growth. Let me now turn it over to Chris to cover some additional market fundamentals on Slide 5.
Thanks, Terry. This quarter, while relatively uneventful from a dispatch and weather perspective, did help validate our driving thesis that data load is here and more is coming. As Terry mentioned prior, below the surface, incremental load is beginning to manifest. We are seeing relative price strength in the face of very benign weather conditions. As cash markets are starting to feel this tightening, so too are the forwards. Looking at the chart on the right, you can see the market response over the past couple of weeks. And as you can see further, it has been a recent phenomenon, but one that we have been anticipating. Power is up, sparks are widening, and it remains a good time to be in power. And further, a really good time to be an IPP focused on being an IPP. Our native position is long several thousand megawatts of generation spread across the supply stack. Outside of the PTC and the data PPA, we have no other natural hedges. This informs our commercial strategy and allows us to participate meaningfully when we take a view on the market. And you can reference the appendix in back for those specific percentages. Similar to the forward, PJM capacity markets have been reflective of these tightening fundamentals as well, expressed through the all-time high BRA clears. PJM shrinking reserve margins continue to be a topic of discussion, both inside and outside of the RTO. PJM and the DOE have flagged potential supply shortfalls by 2030 if the trend isn't reversed. In addition to needing new supply resources, PJM's existing asset base will have to be relied on heavily to ensure grid reliability moving forward. The average amount of uncleared megawatts in the last two auctions was less than 1 gigawatt, and the recent released '27, '28 auction parameters show further evidence of continued tightening fundamentals. This auction is the final auction with capital floor in place. I'll note, we remain active participants in ongoing stakeholder discussions with PJM and other governing bodies, and we will see the results from that capacity auction on December 17. Back to you, Terry.
Now to Slide 6, which covers our year-to-date financial and operating results. For the nine months ended 2025, we are reporting $653 million of adjusted EBITDA and $232 million of adjusted free cash flow. Our liquidity remains substantial with $1.2 billion of liquidity available for working capital, including approximately $490 million of cash available. Once we close on the Freedom and Guernsey acquisitions, we'll have $200 million more of liquidity as our revolver capacity will increase to $900 million. Excluding the acquisition financing, our leverage ratio is still within our 3.5x net debt to adjusted EBITDA target. Year-to-date, we generated 28 terawatt hours with over 40% of this generation coming from our carbon-free Susquehanna nuclear facility. Our year-to-date forced outage rate is higher than we have experienced in the past. This higher outage rate was largely driven by outages at our Martins Creek plant, which experienced prolonged outages due to induction fan repairs. These issues have been resolved, and the plant continues to serve as a peaking unit across the fleet. However, the outages did contribute to our inability to capture some upside as previously noted. Safety remains our first priority across the fleet, and our year-to-date recordable incident rate was 0.64. While higher than prior quarters, this remains well below the industry average. The commitment of the team to operate in a safe and reliable manner is an important part of Talen's value proposition. Now turning to the financial results on Slide 7. For the third quarter 2025, Talen reported adjusted EBITDA of $363 million and an adjusted free cash flow of $223 million. This quarter, our earnings include the higher 2025, 2026 PJM capacity pricing of approximately $270 per megawatt day and an increase in energy margin. Adjusted free cash flow includes higher CapEx associated with the extended Susquehanna refueling outage. Additionally, we also have higher solar energy pricing, which resulted in increase in generation across the fleet. Moving now to guidance on Slide 8. With three quarters behind us, we are narrowing our 2025 adjusted EBITDA as we are trending towards the low end of guidance, as mentioned at our Investor Day, due to the lack of volatility in prices in the third quarter and the extended outage at Susquehanna that offset our strong first half of 2025. Adjusted free cash flow remains near the middle of our original range, driven by our continued focus on generating the most cash flow per share as possible. We are affirming our 2026 guidance and remain confident in both our adjusted EBITDA and adjusted free cash flow numbers. All of this remains consistent with our Investor Day. Turning now to Slide 9. We remain committed to returning capital to shareholders. In September, we announced another upsizing of our share repurchase program, and we will have $2 billion of capacity remaining through year-end 2028 once we close on the acquisitions. We are supportive of targeting $500 million of annual share repurchases during the post-acquisition deleveraging period. Once we reach our targeted leverage of 3.5x or less, we intend to return to allocating 70% of adjusted free cash flow to shareholders on a significantly higher free cash flow base. Lastly, during the quarter, we sold nuclear PTCs for approximately $190 million. We were able to monetize the credits due to the additional benefits from tax reform implemented this summer, along with the tax benefits from the upcoming acquisitions, which will significantly reduce our cash tax burden for the next several years. The monetization is just another example of Talen's focus on producing bottom line cash flow. The liquidity addition from these sales provides us more options on our deleveraging activity, share repurchases and other strategic transactions. Turning to Slide 10. As of October 31, our forecasted 2025 year-end net leverage ratio is approximately 2.6x, well below our target. We will be focusing on debt paydown after the acquisitions to reach our targeted net leverage ratio by the end of 2026. And our pro forma net leverage is expected to remain below 3.5x by year-end 2026. With that, I'll hand the discussion back to Mac.
Great. Thanks, Terry, and thanks for everyone for joining us on the call. We look forward to your questions. We'll turn it back to the operator, Michelle, and open the lines for questions.
Questions and answers
Operator instructions were provided. And our first question is going to come from Angie Storozynski with Seaport.
So my question is, as you said, Mac, every day, we seem to be getting announcements about new power deals, just not from IPPs, it seems. I mean we have conversions of Bitcoin miners, oil and gas companies are jumping in, companies that completely had nothing to do with power until yesterday, seemingly building new nooks and gas plants and yet we're still waiting for public and private IPPs to monetize their assets. So you're probably the last person I should push back against because you have been doing your share and some. But are you concerned that existing assets are losing the time to power benefit to the new build and those conversions?
Angie, I appreciate the push. No, we're not concerned. We're going about our business, and we have been. As I said during some of the remarks at the opening, we've continued to work to execute. One of the things we've done over time is learn a lot of commercial lessons. We still think that's very applicable, and we still think we offer speed-to-market solutions. These things are complicated. They will come; they just take time. As I said during my comments, we want to do the right deals on our timeline and with counterparties whose timelines align with ours so we can find the appropriate point to execute. That doesn't change the overall thesis or our thought process around this; it just means things take time. If you think about it, we've gone from a behind-the-meter arrangement — we all know the history — with the ISA being thrown in the air, rejected, and then worked on and solved, and then we went to work on a commercial solution and moved it to the front of the meter, gaining a lot of commercial learning along the way. We've always said we think that gas is the capability, a gas portfolio to be specific, not just single gas units. Using a portfolio like we have gives us the ability to backstop front-of-the-meter transactions that are coming. We continue to make progress. I mentioned, for example, and I'll let Cole jump in here in a second, that we're advancing things. You saw what's going on at the Montour site with the rezoning. We think that's a good opportunity. We've described that in probably too much detail over the past six months so that everybody wants to know when we're going to announce a deal there, but we're moving that forward in a positive fashion. We're going to get there. These things just take time. Cole?
Yes. I would just add that I think the deals that we've done, Angie, and continue to focus on our advantage from a speed to market, and we do provide an advantage there and probably on a different time frame than some of the deals that you're referring to being announced recently. And so we're really focused on those kind of deals that can get sites up and running and powered in the near term, like in our next year, 2-, 3-year window. And as Mac said, they're a little complicated, and we like to go for the gigawatt scale as well. And so it just takes some time, but we're pleased with where we're at and where we're progressing and where we're going.
Good. And just one follow-up. Mac, you mentioned expansion of your portfolio. So I mean, I know that you have the 3.5x net debt-to-EBITDA leverage limit. I mean, is that a real limit? Or could you, for example, address it by, I don't know, securitization of the revenues that are coming from the Susquehanna contract, just being a little bit more creative to give yourself more of a balance sheet mom?
I'll let Terry jump in after he kicks me under the table because I'm always saying managing the balance sheet and credit is his responsibility, not mine. That's a joke, Angie, and for anyone on the credit side, Terry is now literally kicking me under the table. On the 3.5x, as we've said before, the way we do capital allocation, the way we think about hedging and cash-flow hedging, and how that ties into ensuring appropriate cash flows all connect to our overall strategy. We set net leverage at 3.5x and said we were willing to toggle it. We are toggling it for a short period as we take on this new debt and close the Freedom and Guernsey acquisitions, but we are committed to returning to 3.5x net leverage by the end of 2026. We all pull on those strengths. When you look at appetite, as Terry mentioned, there was demand for our paper in this financing; it was well subscribed and we achieved rates better than we anticipated, which is great. There is the ability to pursue options, but on creative securitizations and project-level financings, we have spent several years making a concerted effort to clean up the balance sheet by removing project-level debt and sponsors and bringing obligations onto the corporate balance sheet so we can manage across the portfolio. We believe that having a portfolio and the commercial expertise to structure long-dated, large contracts, together with corporate-level debt, allows us to backstop across the entire fleet rather than isolating credit in project-finance entities that securitization creates. People, and Terry and Sergio can speak to this as well, come to us wanting to put a contract and an asset into a securitized structure, but that approach forces you to quarter off credit and creates ordering constraints when dealing with PJM. Those factors make less sense to us than maintaining a portfolio that can provide long-term contracting solutions.
Yes. Angie, this is Terry. Just to add on to Mac's comments, a couple of things. The 3.5x leverage ratio, it's a target. For the right opportunity with the right return, we would be willing to push past that. The other thing to elaborate on is as we think about serving our debt, serving our interest and just serving the operations of the business. As long as we've got contracted cash flows that have limited risk, we feel really comfortable around that. And this goes back to Mac's other comment is when we think about the balance sheet and we think about the leverage with respect to, okay, how are we hedged, how comfortable do we feel about those cash flows? And how does that sort of near-term outlook sort of all work together. And so we'll continue to do that. The other thing I'll add to it as well is as we've grown the business and as we've added on the initial AWS transaction and the second one, and then hopefully, here in short order, the Freedom and Guernsey assets, we're growing the earnings base. We're growing the cash flow base, which obviously has resonated with the credit market. And so when I got here 2.5 years ago, we had just issued some senior secured notes at like 8.625%. And here, we just get off the back of issuing two sets of unsecured notes that have a 6 handle on the interest rate. And so our cost of debt is going down in recognition of how well the business is performing and how we're growing. So we think those things are all positive. They give us a lot of options as we think about growing the business, and we'll utilize it as we move forward.
And that balance sheet will get stronger over time as we grow into the contracted business and the ramp with the existing AWS contract progresses. Adding more contracts will further strengthen the balance sheet and give us clear visibility into those cash flows. Maybe later, but not now. We like our current position, and it gives us the flexibility to make short-term adjustments and then return to our net leverage target of 3.5 times net debt to EBITDA.
And the next question comes from Shar Pourreza with Wells Fargo.
So, Mac, just in Maryland, there's obviously a vocal IPP around supporting solutions to add incremental capacity in the state's expedited CPCN solicitation process. I guess what's your thoughts around offering up alternatives and maybe again, your view on the construct in that state and then you can kind of be supportive of RA efforts there?
Thanks, Shar. First, we're doing our part through the RMRs. Those are units that were slated to shut down under agreements with environmental firms, and we've worked hard to execute those RMRs. We're evaluating whether it makes sense to bring more gas to that site to prolong operations and convert the unit, but you have to be able to get gas to the site. We're working with the local utility, but that will take time. Given where our assets are located, it probably doesn't support redevelopment at the scale you could do farther east in the Bay toward Chalk Point, where there are a couple of CCGTs and open land. Getting gas across the river or across the Lower or Upper Bay is a difficult proposition. We're working to secure incremental gas to see if we can take that coal unit and convert it like we converted Brandon and Montour and as we can at Brunner. If there were another option, we'd consider it, but right now that's how we see our contribution to Maryland.
Got it. Okay. That's perfect. And then just lastly, shifting gears to Pennsylvania. I guess, Mac, what do you need to see to build there? I mean, have you had any discussions with the utilities on their views of resource adequacy solution? And is there a common ground there that you can strike? It just seems like when you're hearing from Exelon and PPL, there's discussions to be had, but I just want to get a sense from you where the bid ask is there.
Sure. First of all, there's a big debate about how to solve this, and we've been active in it. We're working on a couple of solutions. We developed a proposal with a consortium that included hyperscalers and other IPPs, trying to be constructive. The real issue is the rate shock from capacity prices spiking to $330 in the last clearing and $270 before that. Those levels do not support new builds. If you look at economic analysis from the quadrennial review, you need about $500 a megawatt-day. Then you get into whether one clearing price will incentivize new build. We think structural changes are needed to provide a longer-term revenue perspective for new builds. There is talk about building in rate base, but that only happens when the right contract, ability, or returns exist, and even then it’s around $2,500 per kW. Even though energy curves and capacity markets have ticked up, they still do not provide the revenues necessary for a new-build CCGT. We believe CCGTs will be needed in the future as overall energy demand rises, but today there is plenty of energy on the grid. We can run mid-merit and peaking units more often, and we’re setting our units up to do that. For example, at Martins Creek we are running them harder with less downtime for maintenance. As Terry mentioned earlier, we are making adjustments in CapEx and O&M, which can absorb a lot of incremental energy demand. But the central problem remains solving the 50 to 100 hours per year of capacity shortfall. We think that is better addressed by batteries or peaker plants rather than CCGTs, at lower overall cost, and those resources should be utilized. To deploy them, many people are waiting for the right commercial terms. We can proceed if someone offers an appropriate return via a long-term contract, but energy and capacity markets haven't reached that point yet. I do think you’ll see markets move toward that over the next year or two, but the near-term solution will likely be something other than CCGTs because of speed to market—you cannot build a CCGT quickly enough to meet the near-term need. Many of the alternative solutions people propose are high-cost because of the technologies involved, but they’re being pursued for speed to market. We need CCGTs in the long term, but we must solve the 50 to 100-hour shortfall in the near term. We believe we can participate by adding peakers and batteries to our fleet if the right return mechanisms are available or if we can integrate them into contracts appropriately. I hope that answers your question, Shar.
It does. It does. Yes, there's obviously some more wood to chop there, but I appreciate it.
And the next question will come from Bill Appicelli with UBS.
Just wanted to build upon some comments you just made there. You mentioned earlier about energy prices moving up. What are your thoughts on what's driving that? And where can that go, right? I mean if you discuss what you just mentioned there around the mid-merit and the peakers running more, right? I mean that's going to drive up the marginal cost on the energy side a bit. So I mean, is it driving to the upper $80s, $90 in terms of cost of new entry on an energy level? Or how do we think about sort of what the backdrop is for the wholesale power markets?
Yes. Good question. Let's see if I cannot mix metrics here. Look, on a megawatt hour basis all in, I think you're looking at 120 plus for CCGT on a greenfield development and you're looking at a 2030 plus delivery time frame of COD. I do think that you can find ways to add so-called additionality to the grid that are cheaper than that and quicker than that. And those I've already mentioned. I do think that, and to your point, and this is a little bit about what Chris mentioned when he talked about the curve and the recent phenomenon of going up the curve, some of that's gas driven a little bit, but some of it's also people looking at it and saying sparks are expanding because you're getting to higher cost units filling the energy demand. This overall energy demand is going up. So energy prices should go up. As Chris said, we've been waiting to see that and people often ask us. Even until recently, if you look at that chart that Chris was talking about, it's only been like the last three weeks that this thing has gotten a bid. But we were anticipating that it was going to go there. We set the portfolio up on that to think about when we go in and layer cash flow hedges. Obviously, there's some timing that you can apply there, but there's also, as Terry mentioned, managing the cash flow hedges to protect the downside as we look at security for having cash flows for debt service, share buybacks, and the like. So we do think that energy markets are going to go up. I don't know if it's 80, 85, but we're starting to see on-peaks go up. And quite frankly, for the first time in a decade, we're starting to see summer on peak start to beat winter on peak. That hasn't been a phenomenon for a decade in PJM. And, as I think Terry mentioned, those were the third and fourth highest peak days in June that we've seen. So a lot of this stuff is starting to manifest itself. I think the tip of the iceberg was the 270 and then the 330, and that's the capacity market because it's further out in time. But now you're starting to see it in the '26 and '27 forwards. And we just think that that thesis is continuing.
Yes. And maybe, Bill, to add to Mac's comments, really your second part of your question, I think what you're seeing in the forwards is really driven by what we're seeing in the fundamentals, right? We've seen this demand growth. We've gone, I mean, effectively a decade with sort of flat to no growth in the PJM market from a load standpoint. And now we're starting to see tangible load growth on a year-over-year basis, and that expectation is going to continue. I think the recent peaks that we hit during the summer are other good indicators of it. And then also just people rolling forward and actually really thinking about how they're going to handle their wholesale power cost in '27 and '28 and beyond. So we've seen more activity in the market. We do think it's fundamental based, and we think it's constructive. And to Mac's earlier point, to get to the level of value to where you can build a CCGT or anything else, right, the devil is in the details of what is that total cost and then what return are you solving for. But we're still far away from that point. To Mac's comment, right, we think that, that number is well over $100 per megawatt. And so we've got quite a bit of ways to go before we can hit that point.
Okay. All right. No, that's very helpful. And then as far as additional asset acquisitions or can you just speak to what's out there? And are there things that you would still consider folding into the portfolio at this point when you kind of are making the evaluations around capital allocation?
Well, I think on that, we'll continue what we always say is, obviously, we're always in the market looking for things. When we find something that we like and we think fits where we want to go, we'll definitely let the market know about it. As you've seen, generally, right, there's still a lot of M&A activity in the space. It varies from one asset to small portfolios. Obviously, there's been several larger M&A deals done this year across the IPP space. But still a lot of activity. I think there's still a lot of holders of assets that have held them for a while that are probably in a spot where they're looking for an exit. And then obviously, you've seen Talen and others engage from a buyer standpoint. So, still active, still looking at things as they come around across our desk. And as we get to a place where we find something, we'll talk about it then.
And the next question comes from Jeremy Tonet with JPMorgan.
Just wanted to pivot to battery storage, if we could. And wondering if you could talk a bit about why you chose the partner you chose. And do you have any thoughts specifically on long-duration energy storage for AI data center applications?
Yes. First of all, we're working with EOS and looking at the deployment there. The technology supports long-duration storage but can also be discharged for four hours or up to 12 hours. You can operate across that spectrum, dispatch it over time and reload the battery, which is critical for managing load. When we think about batteries, we think of them as reducing net load and cutting peaks, which ties back to the question of 50 to 100 hours and whether you can do that across time. So we see a good matchup there. I will also say that because of the technology EOS uses, its fire protection requirements are different from lithium-ion or other batteries. Because it doesn't have the same components, it doesn’t need the same fire protection, and that provides an advantage. As we think about load and colocation — placing batteries next to data centers or next to a nuclear unit or one of our other generating units — you don't want to worry about fires and the associated risks, so that is an incremental benefit in addition to the ability to match duration at different intervals. Obviously, efficiencies decline depending on what you choose, but those are the two main benefits I see. And again, that ties back to colocation. Cole, I don't know if you wanted to add anything there.
I think it's an interesting technology and opportunity for us that we're exploring, and better understanding the value the battery solution may provide to data centers is certainly core to why we want to proceed with some kind of partnership. We're exploring it and having discussions around what that value stream is. As those solutions present themselves, I'm sure we'll be talking more about that.
Got it. And just wanted to pick up, I guess, with the forward curve moving up a bit here, as you touched on before. I'm just wondering, granted it's recent, the moves you're talking about here, has that started to, I guess, filter into conversations on long-term contracts, just these upward moves, has that changed anything in conversations?
Power is not getting any cheaper. So I think it informs things. But when we think about long-term contracts, I'm looking at Cole here to jump in. We're thinking about things over the long term. And we like matching that load over the long term, as Cole said, in the gigawatt size and the large size over time. And when you think about long-term contracts, the markets have gone up, they've gone down, et cetera. We like the ability to take our assets, contract them provide the appropriate level of return, reduce risk, therefore, we would contend would generate effectively the ability to have a lower free cash flow yield because it's lower risk and higher valuation. And that's the Talen flywheel that we're trying to implement. But...
Yes. And I think as energy prices go up, it informs us on the terms, not just the pricing, but the overall terms, as Mac is talking about of any future deals. And also, I'm sure, informs the counterparties on what their views are and what term and pricing and risk and so forth. So, look, I mean, I think as Mac said, we're going to do deals on our time line. And I think as power prices are continuing to push higher, it continues to give us conviction that we're going to look at the right deals in size and pricing and all the terms. And there's no need to go rush to contract, but obviously do them prudently at the right time.
I would add that you can't see a visible curve for 10 years, whether for capacity or energy. So we're sitting back thinking about what appropriate returns look like, how we should think about counterparties, how we think about gas, and how we think about backstopping with the portfolio I manage. All those different things factor in. And then what does that do to our credit capacity, which Terry was talking about earlier in managing our leverage ratio? There was a question earlier about whether you could do different things with your balance sheet. As you grow into having a higher contracted portfolio over time, you could certainly do different things, but we're just not there yet. We think overall we've got to embed the right risk and the right premium for taking these longer-term views, and the counterparty has to do the same thing, as Cole was saying, and it takes two to tango.
And our next question will come from Steve Fleishman with Wolfe Research.
Maybe following Angie's initial question, in terms of customer interest, how much is the desire for additionality affecting demand for more contracts for you? And do you see the Sus deal potentially adding SMRs, and do you expect new investment to be part of any deals you do?
Steve, it's Mac. Team can jump in. Look, I definitely think that the so-called additionality — I'm not even sure there's a standardized definition of that — but incremental megawatts to serve incremental load is kind of how I'm going to frame it. I do think that over time, we're not out of energy right now. We're actually not capacity short. Markets are clearing at the administrative caps, and we clear closer to the real caps if left uncapped. But we're starting to get to the point, and these are in the out years, of needing megawatts, and there's also the ramp of data centers over time. At some point those two things are going to intersect and you'll have to ask how we make sure, and it's the so-called BYOP or BYOG — bring your own power, bring your own generation, et cetera. Those things will intersect later in the decade. We're supportive of thinking about how to solve that in a market construct and in our contracting strategies. It just hasn't ripened yet. There's been a lot of talk and a lot of people offering solutions around power without having offtakes associated with them yet. That just hasn't ripened. I don't think — and Cole can jump in here — that the demand or the desire to quickly connect power or to secure power toward the end of the decade is changing. It's just a question of how you solve that. We're actively thinking about it.
Yes. I would say that the data center hyperscalers and everyone involved in data center build-outs are thinking about the problem on multiple fronts. As Mac mentioned before, there are different phases of power: near, medium, and longer term. In the medium and long term, hyperscalers are looking at how to add megawatts to the grid as they ramp up. But they are also very focused on the near term, 2026, 2027, and 2028. We see them actively securing and connecting to existing power, and that is where we are focused. If there is an opportunity to participate in a longer-term solution, such as small modular reactors or new builds, they are having those conversations and exploring them as well. Right now they are concentrated on getting power; many hyperscaler CEOs are talking about doubling capacity by 2027 from 2025, and the only way to do that is through existing grid power. From all of our conversations, they remain very interested.
Okay. PPL also reported today, and their high-probability gigawatts of data center demand by the end of the decade is roughly double or triple their current peak. I'm curious, Chris, could you talk about whether you are seeing the zonal discounts you historically had begin to disappear in the forward curve? At some point, could this flip to a premium? I'm just curious what you're seeing.
Yes, Steve, it's Mac. I'll take this. We lost Chris, unfortunately, due to a conflict, but that is the next evolution of what we expect to happen. If you think about ramp rates and where generation is on the grid today and the basis differential, I think what you're asking, if I'm understanding correctly, is about PPL zone versus West Hub because PPL zone trades at a discount. Over time, as you build load there, you start to soak up the incremental megawatts and you move the East-West interface west, so you get closer to the West Hub and the differential begins to close. But that's probably not a near-term phenomenon until these things ramp up. If you think about our Susquehanna 19 and 20, full ramp is 2031 and 2032, so we're four or five years from that. There's no way for us to go out and test the market, but I think by 2032 we would expect PPL to be a lot flatter to West Hub.
And Steve, maybe just to add to Mac's comments, that basis market historically, especially in the term, the term will react if there's large transmission projects and there's transmission projects that are being executed and built upon. But excluding that or excluding those projects that are out there and getting those completed, it does have a bit more of a recency bias of, okay, how is it showing up in the real time? How is that basis showing up as load comes in. And so as we see more of that as the load comes into the PPL zone, I think that will inform the term market a little bit more. And then you'll start seeing a change with respect to that. But it's a little bit more of a nuanced market and the fact that it trades off of fundamentals around transmission and then just what sort of a more recency bias, if you will.
And then just one last quick one. The Martins Creek issue this quarter, and you mentioned you're just running these peakers harder. Just are you looking at needing to ramp up capital spending at all to do anything to those units for next year?
Yes, that was included in our guidance. We're also thinking about this, Steve. As power markets come off you have to curtail capital plans and rethink how plants are dispatched, and when markets reverse you have to do the same in the opposite direction. We plan to put more capital into 2025 and into 2026. What we're noticing, and I've mentioned before, is that peakers like Montour and Martins Creek are getting extended run times and longer dispatch. Martins Creek was under a 4% capacity factor a couple of years ago and now is running in the teens. In the past we could take units offline for extended maintenance periods; now they're running for long durations and we don't have those downtime windows. So it isn't just additional capital — it's also how we think about maintenance while the units are operating. That was not the underlying issue at Martins Creek this time. We had an ID fan with bearing sets that took a while to obtain, but we didn't have the downtime to replace it. We've been transparent about that. Overall, we're excited because this helps explain why forwards are ticking up. We're operating further up the dispatch curve; every hour is moving higher on the curve, which means higher-heat-rate units are being called, and we're seeing that across our fleet.
And our next question will come from Nicholas Campanella with Barclays.
So just on the commentary about Amazon doubling their overall capacity by '27, that's their number. But just can you talk to the AWS ramp? Obviously, like that data point is supportive, but just any indication that they could be ramping power draw sooner than expected? Any data points on the ground level that you can kind of point to?
Yes. So Nick, maybe just to touch on that and then I have Cole chime in on this as well. As Mac mentioned in the opening remarks, obviously, the data center is powered, it's energized. They're taking electrons and moving forward. The construction of the site continues in earnest, significant amount of work done over the past several months, and they continue to push forward. Obviously, we don't want to get into too much detail around what their ramp is like and what they're doing from a confidentiality standpoint. But we're fairly comfortable with how they're pushing forward. Cole, do you want to add anything?
I would reiterate what we've said before about continued progress. Multiple buildings are being built on top of the building we sold to them, with additional ground preparation, and anyone who drives by can see all the activity. One point to remind folks, as we mentioned back in June, is that we transferred the old Nautilus buildings, which can support up to 200 megawatts, to Amazon. What they do with that is for them to speak to, but we see many signs that acceleration will continue and that they will ramp up faster than the timelines we published in June.
Awesome. And then just one cleanup question about the acquisition. I hear in your prepared remarks that you're working constructively with the DOJ and could target closing sooner than first quarter 2026 now, but 2026 guidance assumes a January 1 close. Can you talk about the conservatism in that 2026 number and your ability to maintain that range if the close gets pushed out?
Yes. There are a lot of moving pieces to a guidance range. Forwards are up, and that's within the range we provided. There's potential for delay; as we stated, this might slip into Q1. We're still hopeful and working hard to get it done this year, but we'll see what happens. As we get closer to that time frame, we'll provide updates. We're not at a point that suggests we should change the range right now. As we get closer, we'll give you an update.
And then our next question will come from David Arcaro with Morgan Stanley.
I was just curious maybe where are we in terms of economics for battery storage in PJM, just thinking about that EOS partnership. And wondering if you could characterize how you're seeing that opportunity across your fleet if you were to add battery storage.
Yes. We're not quite there, David. Sorry for using your name. Look, I think that we're not quite there and ready to get into that. I think we're in the early days. But as we look at the economics and the capital build, and as the cost of batteries comes down and their ability to basically think about it, it goes back to, I'll call it, load balancing. For all practical purposes, you could think of it as supply balancing: we've got a lot of incremental megawatts that can be run most hours except the peak. So why not use those to charge batteries and then discharge them during the peak? We think that can become a more economic piece over time. Obviously, batteries need to continue to evolve, bring their capital cost down, and then bring their optimization methodologies into play. We think that works well when you think about how to take a load that is, for the most part, a base load — I'm talking about data centers — and clip some of the peaks of that, and then all you're doing is using all the rest of the hours other than the 50 to 100. So we think it's going to show some promise, but we're in the early days of getting through the economic model. So it's just, David, it's just too soon.
And the next question will come from Julien Dumoulin-Smith with Jefferies.
Look, let me clean up a couple of things here. One, on buybacks. I noticed there wasn't much activity this quarter. Do you want to speak to that real quickly? And perhaps more relevant, again, you're going to laugh here, Mac, but MNPI, anything to talk to on that front? Is there any twist there that's relevant?
Yes. So, Julien, with respect to the buybacks, obviously, we were fairly active during the quarter with the Freedom and Guernsey acquisition and then also getting ready for earnings and financing. So that really drove sort of the buybacks with respect to Q3, working hard to get those acquisitions done and announced and out there and then working on our Investor Day materials as well. So that was a big driver of where we stood on buybacks in Q3.
Okay. Sorry, you guys did do buybacks in Q3. That's what I was confused by.
No, we did not.
We did not. And you can see that we didn't disclose it, obviously, Jordan. As Terry said, there's a lot going on during the quarter, whether that's MNPI, getting ready for financings and blackout periods around the results, or closing Freedom and signing Guernsey. You can call that MNPI in arrears, maybe, and that had a large preclusive effect across the quarter.
Absolutely. A couple of cleanup items. Is there any chance that you accelerate this Sus deal with AWS and then clean up the remainder of the uncontracted Sus here? I mean I just want to come back to that concept. I mean you talked about the ramp-up to 2031, 2032. Everyone is grappling to get this stuff faster. In theory, this is a question of execution, right, to get it done faster with you guys. Can you speak to that a little bit on the ability to ramp it up?
Look, anecdotally, we talk about desires regarding Amazon. Concerning the contract, we need to maintain confidentiality with our counterparty. We can say anecdotally that they are speeding things up. We said it's a lever that can be pulled. We stand ready to deliver 19 to 20 megawatts whenever they want to take it. That was the obligation we stated. If you think about speed to market, when you have a site under construction and megawatts ready to go, there is nothing that accelerates speed to market more than that. That's about all we can say because we won't speak for Amazon; you would have to talk to them. We're excited about the possibility of them ramping up, and if they did, we're willing and able to serve.
Excellent. All right. No, fair enough. I appreciate that. And then lastly, just would you be in a position to get another gas contract here prior to the close of these acquisitions? Or would you tie those two together to the extent to which you get the acquisition gets kicked out in 1Q?
Look, I think you've written about this, frankly, so it's a good thing to have a discussion. When we thought about this, let's step back to before Freedom and Guernsey. Before that announcement, we were already working on the next deal, and that was even before we revamped the 19, 20 front of the meter — it's just a continuation of the process. When we looked at it, we noted we have Montour and Martins Creek and still have 300 megawatts available there, so I don't see the intersection of those two as a reality. Obviously, with the Freedom and Guernsey acquisition, if we do a deal we would eat into the megawatts we currently have, and therefore Freedom and Guernsey reloads the bank. That's how we view it.
That concludes our allotted time for questions. I would now like to turn the call back over to Mac for closing remarks.
Yes. So, thank you, Michelle, and thanks, everyone, for joining us and for the Q&A period. We tried to spend some extra time, and I know there's a couple of people still in the queue that we didn't get to. We're happy to take your questions and look forward to seeing everybody in the balance of this year and early next year with a busy schedule on the road at conferences. And I appreciate everybody's interest in Talen, and we look forward to continuing to execute. Have a great day.
This concludes today's conference call. Thank you for participating. You may now disconnect.