Ladies and gentlemen, thank you for standing by. Welcome to Talen Energy Corporation investor update call. At this time, all participants are in a listen-only mode. After the speaker's presentation, there will be a question and answer session. To ask a question during this session, you would need to press star one one on your telephone. You would then hear an automated message advise your hand is raised. To withdraw your question, please press star one one again. 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. Sir, please go ahead.
Thank you, Michele. Welcome to Talen Energy Corporation's 2025 Investor Update Conference Call. Speaking today are Chief Executive Officer Mark McFarland, Chief Financial Officer Terry Nutt, Chief Commercial Officer Chris Maurice, and Executive Vice President Strategic Ventures Cole Muller. We posted our investor update presentation this morning to the investor relations section of Talen's 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 the appendix of our investor update presentation. With that, I will now turn the call over to Mark.
Great. Thank you, Sergio, and welcome to this year's investor update call. As always, we appreciate your interest in Talen Energy Corporation. A lot has changed since our investor update a year ago and our emergence just over two years ago, both in the IPP space and for Talen. From a general perspective, we remain steadfast in our view that the intersection of power and data is here, accelerating and driving significant opportunities for those that have the ability to meet customer needs and for those that have the ability to adapt and risk manage solutions under long-term contracts. Since last year's investor update, our market cap has more than doubled, so the logical question is why Talen now? The simple answer is we have a demonstrated history of value creation that we think we can continue to leverage.
We see a strong set of base financial projections with multiple levers to pull to execute the Talen flywheel and create additional value through free cash flow per share growth. Turning to slide four in the presentation, as we have said for the past few years, we are an IPP focused on being an IPP. We are concentrated in PJM, the largest deregulated RTO, and in Pennsylvania in particular. It is a market that we know and like and gives us exposure to increasing power demand fundamentals and data center participation. We are very excited about adding Ohio as we see this as an existing untapped market to serve large loads. As all of you know, we measure value creation as adjusted free cash flow per share and adjusted free cash flow per share growth.
To be clear, that is adjusted free cash flow after taxes and after growth CapEx, effectively it's cash flow available for distribution. The value creation to date that has been achieved by focusing on execution has allowed us to repurchase $2 billion in shares, or roughly 23% of our shares outstanding, and we plan to continue repurchasing shares going forward. It is worth noting that these share repurchases have helped create over $3 billion in value on a mark-to-market basis. We were the first in the IPP space to announce a data center deal with Amazon Web Services, with a behind-the-meter deal in 2024, and at the same time, we sold them the campus for $650 million. We then leveraged this first mover advantage to close on a larger and revamped front-of-the-meter deal with Amazon Web Services earlier this year.
For those that are new to the story, that is an $18 billion notional contract lasting over 17 years with options to extend even further. This contract resulted in visible, stable, and growing free cash flow per share, and that is for not just one year, but for many years to come. With our strengthened balance sheet, we were then able to strategically and accretively sign deals to acquire the Freedom and Guernsey plants, which will add approximately 3 gigawatts of long-term contracting capability, as well as diversify our fleet and expand our opportunity set to Ohio, as I previously mentioned. Upon closing the acquisitions, we will generate over $3 billion of cash through 2028, and we are targeting $2 billion available for shareholder returns between now and the end of 2028, while maintaining net leverage at less than 3.5 times.
Given our strong cash flow projections and balance sheet, today we are announcing that the board has approved increasing our share repurchase program to $2 billion, and we have extended the term of this program through 2028. This is just the start. We have additional upside and value creation opportunities by executing on our Talen flywheel strategy. This means contracting large-scale data center customers to further generate visible, stable, and growing cash flow. We can accelerate volumes in our current contract, or we can add additional long-term contracting opportunities, including a remainder of 300 megawatts of capacity at Susquehanna in addition to our recently expanded gas fleet. This leads to an even stronger balance sheet that allows us to toggle between adjusted free cash flow per share growth via share repurchases or accretive M&A opportunities.
The cycle repeats itself as we look to further contract out the new capacity while preserving flexibility for our capital allocation. In a couple of slides, we will unpack how this delivers both a strong base set of projections with multiple levers for additional upside. We turn to slide five. We've done a lot since we emerged a little more than two years ago. As I mentioned, we've now signed two data center deals, culminating in nearly 2 gigawatts under contract with Amazon Web Services. We're in the process of closing 3,000 megawatts of highly accretive combined-cycle gas turbines that I mentioned and adding to our contracting strategy ability. We've been disciplined with our cash. I mentioned buying back the $2 billion in shares to date, but as you can see on this chart, that isn't all we have done. We have sold assets and recycled the capital.
We cleaned up the balance sheet, removing project financings and lowering interest rates. We removed project partners and redeemed equity warrants. We bought out the coin business and then used that to expand the AWS contract, all value-creating transactions. We also went back to the basics in operations and focused on staying low cost. We also reduced collateral and credit costs through our efforts. We listed on NASDAQ last year. It's also worth noting that by executing and therefore increasing our market cap, we have been added to several major indices and most recently were added to the S&P 400. Let's talk about the future on slide six. First, a couple of things about this slide that help lay out our projections. In the second column, we provide overall drivers per year by year. In the third column, our capacity pricing by year.
In the fourth and fifth columns, we provide observable power and natural gas prices using West Hub and Tetco M3. These prices are on a calendar basis for 2026 through 2028, with the exception of 2025, which is actually the balance of the year. These prices are all as of July 31 of this year. Turning to each of the drivers by year, we now expect our 2025 guidance to be at the lower end of our guidance range. In the first half of the year, we extended the outage at Susquehanna, which we felt was the right thing to do, and we are seeing the megawatts that we expected to recover. However, it came at a cost and lost opportunity margin. That was largely offset by a strong start of the year and a strong June, leaving us with line of sight to our guidance range.
All that said, summer has been soft. Overall price formation and volatility was muted, even with high loads, and gas continued to come off as well. We weren't able to realize the upside we typically do in the summer. This is all temporal in our view. Average and peak loads continue to be strong, and as all of us in the power markets know, a weak current year leads to recency bias and drags down forward years. That is what you can see on the right side of this chart as 2026 through 2028, energy and sparks remain flat on a year-over-year basis. Power is flat and gas is flat, therefore, sparks are flat. Chris will talk about this in more detail later and how we are positioning the portfolio, but it simply does not make sense to us.
If prices start to respond as we think they should, it would obviously move us higher in the ranges we provide in the next several slides. Turning to our updated 2026 guidance, 2026 is materially higher than our initial outlook from a year ago and is driven largely by two things: the Freedom and Guernsey acquisition, which we assume to close on 1/1/2026 for these projections, and the tax benefits from recent legislation, which for the most part eliminates our federal tax rate down to roughly 2% to 3% range in the upcoming years, as shown on one of the slides in our appendix. We rolled the new plans and the tax benefits throughout our projections in 2027 and 2028. Note that 2026 incorporates a similar extended outage for Unit 1 at Susquehanna. This past year, the outage for Unit 2 was extended by 40 days.
In 2026, we expect to cut that in half to approximately 20 days incremental for a total outage duration of around 55 days next year on Unit 1. We were able to do this because we have had time to optimize the 2026 schedule and learn from the work we completed this year. Like last time, we expect to realize megawatt gains in the return of capacity. In the appendix, we provide details on both the 2025 and 2026 outages so you can normalize future year projections as we have done, as we expect to return to normal refueling outages in both 2027 and 2028 outlooks. Moving to capacity prices, capacity prices are also materially higher in these projections. The 2026-2027 capacity auction cleared at the cap of $3.29 a megawatt day. Without the cap, the expected clearing price would have been closer to $390 per megawatt day.
For simplicity, we have held the $3.29 a megawatt day clearing price flat for these projections, so effectively for the remainder of our outlook and for the future planning years. Lastly, our growth in adjusted EBITDA and adjusted free cash flow are also underpinned by our Amazon Web Services PPA, with its contractual volumes that started at 120 megawatts in 2025, grows to 240 in 2026, 360 in mid-2027, and then 480 mid-2028. As I said previously, these projections are at contractual minimums. I'm sure there are some who might be confused about the 2027-2028 ramp in our projections, given the wide variety of estimates in those years by the analyst community and investors alike. When we announced the restructured AWS deal in June, we showed only 2026 and then min/max for 2029.
Min being the minimum commitments and max being the max allowable contract volume without agreement or notification between the parties. You should not straight line interpret from our June disclosures between 2026 and 2029, as that would lead you to higher numbers in both 2027 and 2028 than we show on this page. That is not how the contract works for minimums because there is a significant increase in 2029 minimum commits over 2028. It is basically a step function, not linear. Cole will provide some additional detail in his section, but we view any acceleration of the current AWS contract as upside versus these minimums. Let's not lose sight that this contract is not about one year or a couple of years. It is about 17 years with meaningful extensions. It is about stable cash flows with a AA rated credit counterparty.
To us, this contract creates a revenue stream and margin for the Susquehanna plant that should be valued very differently than merchant revenues. In fact, we think in terms of DCF value using discount rates that should trade very tight to treasuries, probably inside of 100 basis points. It strengthens our balance sheet, which provides debt capacity. The credit market has recognized the underwriting benefit of having this set of cash flows in the business, and it continues to support both better debt capacity and pricing for Talen. Additionally, we believe these cash flows should trade at a low free cash flow yield and therefore a higher multiple. That is why we look to continue to execute on our long-term contracting strategy, repeating basically what we have done with this second Amazon contract. This will further strengthen our balance sheet and recycles capital.
On slide seven, we are focused on creating shareholder value, which I mentioned before, we measure by adjusted free cash flow per share and growth of that adjusted free cash flow per share over time. The chart on the left shows Talen's free cash flow per share is expected to grow by 35% through 2028 and is increasingly driven by stable sources as we put more and more megawatts under the contract with Amazon. Again, as we have done in the past, the chart on the left assumes the current share count remains constant over the years. As a reminder, and as I'll say it again, this is adjusted after tax cash flow and after all CapEx, effectively cash flow available for distribution. Versus last year's investor update in 2026, estimate of $15.55 per share at the midpoint, represented by the green triangle or diamond on this chart.
We are now guiding to a midpoint of $23.60 a share for 2026, providing an outlook in 2027 with a midpoint of $27.10 a share, and finally showing a 2028 as $27.40 per share plus. Our 2028 midpoint estimate plus is intended to convey that the Amazon Web Services contract at minimums increases this estimate over 2027, partially offset by increasing cost structure, costs that everyone in the industry is seeing. This 2028 outlook is holding, again, capacity flat for the $3.30, and energy and sparks remain flat on a year-over-year basis. What's important is that we also view the right side of this slide. We show a table with important free cash flow per share growth upsides to years 2028 and beyond. We view these as levers we can pull to create incremental value.
Obviously, these depend on the timing and execution and the order by which they are executed, and there is some overlap depending upon how they get executed. I'd like to take a minute to walk through each of these in detail. First, $2 billion of share repurchase, all done at today's share price, would be about 5 million shares repurchased. This is roughly 10% accretive to our 2028 baseline of $27.40 per share, or an increase of about $3 per share. The chart on the left has shares held constant, and this is simply hitting our allocation that Terry Nutt talks about of 70% of adjusted free cash flow returned to shareholders as part of our capital allocation program.
Second, accelerating a full incremental 480 megawatts of the Amazon Web Services PPA into 2028 over the contractual minimums that we have included in the base projections would add in excess of $100 million in EBITDA and free cash flow, or more than $2 of free cash flow per share growth in 2028. As I said earlier, we have minimums in our projections, so any acceleration of 480 would add to 2028. Next on the list is accretive M&A. We have sized this to be an approximately 1,000 to 2,000 megawatt generic type of opportunity, obviously combined-cycle gas turbines. We think there are additional opportunities out there to do what we have done with Freedom and Guernsey, which are included pro forma in our projections at greater than 50% plus of free cash flow per share accretion in 2026 through 2028.
Here we show 10% to 20% free cash flow per share growth in 2028 for this hypothetical M&A activity. That is because of a couple of things. One, we don't have the same tax appetite given we are not a substantial taxpayer after using the accelerated depreciation from Freedom and Guernsey. Two, to maintain our balance sheet, we may toggle and use some equity consideration in an acquisition, making it not as accretive. Please understand we will only do this if the deal is more accretive than buying our own shares. Three, we are now working off a larger adjusted free cash flow per share base. Lastly, we have sized this opportunity to be between one and two gigawatts, as I previously mentioned, which is about half the size of the Freedom and Guernsey acquisitions, so a little bit smaller.
I am sure that many of you will want to explore the details of what we see out there in terms of deals and how we might structure a deal, but we aren't going to discuss specific M&A details as we haven't in the past and won't in the future. We believe these opportunities do exist, however, and they could add another $2 to $5 per share growth in 2028 and beyond. Last in the table, moving to a new one-gigawatt data center PPA. No one should take this out of context. We standardize this at one gigawatt for simplicity so that you can do the math to increase or decrease the size. It is no secret that we have been working to expand our long-term contracting capability and portfolio and have been doing so since we signed the first PPA in March of 2024.
This one will depend on pricing, ramp rate, et cetera, but we see an additional gig providing approximately 10% to 15% free cash flow per share accretion over 2028 projections, which again translates to $2 to $3 per share of free cash flow growth. When you add all these up, we see potential incremental per share growth of 40% plus. As I mentioned, there are some interdependencies in the table above, but that 40% translates to $11 over the 2028 baseline year of $27.40 per share. As I said earlier, we are proud of what we have created, but we are not done. These are meaningful levers we believe we can pull to further our value creation. One final comment on this slide before handing it off to Terry.
As you'll see in the lower right-hand side of this slide in the blue box, we provide what the Amazon base contract post-2028 would provide over the years 2028 through 2032, 2029 through 2032. As I said, there's a step function from 2028 to 2029. If you just think about the deal off the base 2028 outlook and hold all else constant, and Cole will take you through this, this contract will grow free cash flow per share by roughly 20% from 2029 to 2032 and will continue to convert merchant megawatts to contracted megawatts. With that, I'll turn it over to Terry.
Thank you, Matt, and good morning, everyone. Now to slide eight, where we introduce formal adjusted EBITDA and adjusted free cash flow guidance for 2026 and outlooks on our 2027 and 2028 metrics. 2026 adjusted EBITDA's midpoint is $1.9 billion, an increase of approximately $600 million from our prior 2026 outlook. While the midpoint for 2026 adjusted free cash flow is $1.08 billion, also an increase of approximately $360 million from our prior outlook. Keep in mind that the 2026 ranges include the impacts of the extended outage for Susquehanna Unit 1 that Matt noted earlier, as well as major outages planned for Freedom and Guernsey. For 2027, we provide an outlook with an adjusted EBITDA midpoint of $2.04 billion and an adjusted free cash flow midpoint of $1.24 billion.
For our 2028 outlook, we provide an adjusted EBITDA of $2.06 billion plus and an adjusted free cash flow of $1.25 billion plus. Over the period presented, our adjusted EBITDA is expected to grow by approximately 25% per year through 2028, while adjusted free cash flow is expected to grow by 35% per year over the same period. Our contracted stable sources of margin are growing over time, but we still have meaningful exposure to the wholesale power market and its constructive supply and demand dynamics. Another variable to our projections is the results from capacity auctions. As Matt mentioned, our simplifying assumption is to keep the 2026-2027 capacity auction result flat at $329 per megawatt day for future auctions. As we have done in the past, we provide some sensitivities related to capacity auction outcomes.
A $50 per megawatt day change in the 2027, 2028, and 2028 and 2029 planning year prices result in a $90 million and a $160 million change in adjusted EBITDA for 2027 and 2028, respectively. Continuing our targeted capital allocation of 70% of adjusted free cash flow, these projections allow over $2 billion of cash available to return to shareholders through our recently upsized share repurchase plan. Before passing the discussion over to Chris, I wanted to highlight that we have provided several detailed slides in the appendix to help the investment community with modeling inputs for the numbers I just covered. This detail includes projected generation volumes, capacity factors, forecasted realized power prices, estimated O&M and capital cost, and other items. Specifically, I wanted to point out the assumptions on slide 24 related to our forecasted cash taxes.
As Matt mentioned earlier, we expect to see a significant reduction in our cash taxes for the next several years as a result of the federal tax reform, with 2026 and 2027 cash taxes ranging from 2% to 3% of adjusted EBITDA, respectively. This provides further enhancement to our adjusted free cash flow for the next several years. With that, let me pass it over to Chris to discuss our commercial hedging program and power market fundamentals.
Thank you, Terry, and good morning to everyone. I find myself saying it now more frequently than ever. It is a good time to be in power. Sparks are up, demand is surging, and there continues to be massive amounts of buzz surrounding AI, data centers, national security, and where all that energy is going to come from.
Talen remains at the fore of solving that equation, providing critical energy infrastructure to millions of homes and large data customers alike. Looking at slide 10, national NPGM load forecasts both show dramatic growth expectations for summer peak load. As evidenced by emerging trends this most recent cooling season, the grid was stressed when confronted with any semblance of seasonal demand. 2025 summer peak load hit levels not seen in 15 years, with the heat event in June resulting in the third and fourth highest load days of all time. These forces will continue as the grid is forced to navigate this new flow of megawatts. The recency bias, as a result of the lack of summer volatility, has muted some of the shape that typically begins to form in the outer summers and winters. While sparks have been moving, the forwards are not yet reflective of these growing patterns.
On to slide 11 and looking at forward pricing. Intentionally, there really isn't a whole lot to see here. I included these charts not for filler, but as a reminder that the lack of appreciation in year-over-year power prices provides us with a market opportunity. As one Warren Buffett used to recite, Mr. Market often has volatile mood swings and can act irrationally, creating opportunities for the patient investor. It's not often Mr. Market gives you such compelling opportunities, so we remain anchored in our fundamental analysis, convicted on our constructive use of pricing, and are positioning ourselves accordingly. Turning to slide 12, PJM's shrinking reserve margins continue to be a topic of discussion both inside and outside of the RTO. PJM and the DOE have flagged potential 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. Administratively capped at $329, with uncapped pricing determined to have cleared even higher at $390. The average amount of uncleared megawatts in the last two auctions was less than one gigawatt. Again, further evidence of limited supply available to meet these growing demand patterns. In future years, PJM shows a potential 50% to 100% increase in the cost of new builds. Final values will vary based on the scheduled reviews, but again, any increase to new entering costs would have a resulting impact of shifting the demand curve upward. We will see those results from the 2027, 2028 capacity auction on December 17th. Turning the page.
While this increasing demand growth is promising for future power prices, it is vital that we maintain discipline and rigor around managing the firm's commodity risk. At Talen, we implement a pragmatic hedging strategy that continues to create value through the ups and downs of the commodity cycle. As Matt mentioned, summer 2025 was soft from a pricing perspective, and that has limited some of the forward premium that is typically priced into future contracts. Bound by the limits within our risk policy and in constant coordination with our credit team, we are roughly 50% sold for next year, with nearly a quarter of our generation hedged for 2027. Of note, these hedge percentages fall below our historical ranges for prompt and prompt plus one hedging activity.
Again, reflective of our growing belief that the lack of contango in forward markets will resolve itself through time and more accurately begin to reflect the tightening supply challenges that lie ahead. To give you some perspective, the chart on the right side of the slide provides the margin impact of changes in price, and as stated earlier, our position is skewed long with meaningful uplift in any upward movement spark. With that, I'll turn it over to Cole.
Thanks, Chris. Moving to slide 15, I'd like to provide additional color on our recently announced acquisition of the Freedom and Guernsey plants. We've said it before, the size and baseload nature of these acquisitions are the same as adding another large nuclear plant to our portfolio and also deepens our exposure to Pennsylvania and Ohio, some of the fastest growing data center markets in the country.
These acquisitions stand on their own merit as premium cash flowing assets in the PJM market, and we look forward to closing on these assets in the near term. We also believe that these assets provide additional diversification benefits to our portfolio and give our data center contracting platform valuable flexibility. Freedom and Guernsey add approximately 3 gigawatts of generating capacity from high efficiency CCGTs that are among the newest and lowest heat rate plants in the PJM supply stack, a feature that is increasingly valued in long-term contracts. The plants also benefit from advantaged fuel supply costs due to their close proximity to the Marcellus and Utica shale. In terms of closing timing and approvals, we have filed all required regulatory filings and expect to close these acquisitions by year-end after FERC review and the mandatory HSR waiting period. One procedural note.
Today, we are refiling our HSR submission at DOJ's request for an additional 30 days to consider the submission. I'll emphasize that this is not a second request from DOJ, and we don't expect the extension to impact our closing timeline. Some of you may recall a version of slide 16 from our June 11 business update that focused on our second deal with Amazon Web Services. We have refreshed the slide to show the total cash flow per share with our latest 2026 guidance and our outlook for 2027 and 2028, maintaining the relative impacts of the revised PPA that we shared back in June. We continue to build on the 2028 point estimate outlook that Terry Nutt discussed earlier.
It's noteworthy that there is still significant built-in cash flow per share growth as the PPA ramps to its full potential, up to 20% additional cash flow growth that is supported by simply executing the current PPA beyond 2028. We believe this built-in contracted growth should continue to drive down cash flow yields and provide further increased value per share. We continue to believe there's significant opportunity to pull these benefits to earlier years if and when Amazon Web Services ramps faster than the ramp shown. As a reminder, the contracted ramp schedule through 2028 remains the same as the original PPA, i.e., 120 megawatt ramp per year from 2025 through 2028. Though there is flexibility for Amazon Web Services that's significantly ramped into these volumes above these commitments, we are showing the high end of the volume range at 960 megawatts through 2028.
Though with notice, AWS may even exceed these volumes all the way to the full 1,920 megawatts. Starting in 2029, we show the impacts of the step function increase in minimum commitments to 360 megawatts per year until reaching the full capacity. There is flexibility for continued acceleration. An example of this flexibility includes Amazon's ability to leverage our PPA across all of their Pennsylvania sites as they develop in parallel. Just one example of contracting flexibility that helps both Talen and Amazon and demonstrates our ability to work constructively with counterparties to drive incremental value. As a reminder, at the minimum ramp schedule, this contract has a notional value of approximately $18 billion over the 17-year PPA that locks in a significant premium and provides valuable stability to our cash flows that we believe provides yet another differentiating factor for investors. Turning to slide 17.
As we continue to execute on our flywheel strategy to power the future, we believe our team has developed a platform with the right mix of assets, technical know-how, and commercial contracting creativity to drive further growth. Hyperscaler needs continue to evolve, and we believe our platform is advantaged to meet what we see as four key hyperscaler priorities. First, speed to market through access to reliable power and critical infrastructure, features often found around large generation assets located in regions and zones with excess supply and transmission, sites we have across our portfolio and continue to advance. Second, gigawatt plus scale sites, sites in growing data center pockets that can expand to multiple gigawatts, either on-site or within a clustering radius. We believe our sites fit nicely in attractive data center growth markets to meet this scale. Third, long-term contracting capability across key Ts and Cs.
For example, price certainty, load flexibility, being capacity backed through physical assets in proximity, and getting protection with appropriate credit support. As we have demonstrated in our first two deals, we can tailor deals around these elements along with managing the appropriate risk through creative arrangements. Fourth, lowest carbon power sources, which includes an increasing focus on best-in-class heat rates from premium CCGTs. Once again, assets we have and will soon expand on with the closing of the Freedom and Guernsey acquisitions. We look forward to continuing our track record of innovation with our customers and providing new and creative first-of-its-kind solutions as needs continue to evolve. With that, I'll turn it back to Matt.
Great. Thanks, Cole. There's been lots of positive momentum since our last investor update in 2024. Today, we laid out a strong set of financial projections incorporating the latest AWS deal, the pending Freedom and Guernsey acquisitions, and the recent tax benefits, as well as other knowns. We have more than doubled our market cap since last year because of these activities, but we are not done. As we have laid out, we see meaningful free cash flow growth from multiple levers deriving from the Talen flywheel, and we look forward to powering the future. Thank you for joining us on the call today. We will now turn it back to the operator and open the line for questions.
Thank you. As a reminder, to ask a question, please press star one one on your telephone and wait for your name to be announced. To withdraw your question, please press star one one again. Our first question will come from Angie Sturzynski with Seaport. Your line is open.
Thank you. Where do I start? Maybe you know you promised us an exciting 2025. You've definitely delivered already, but I'm just wondering if there is more excitement to come before the end of the year. I'm not asking about M&A, just mostly about contracting of other assets.
Look, Angie, it's a great question and I appreciate acknowledging what we've done to date. We're never resting. I think we're always looking at both how do we advance our contracting strategy as well as our long-term contracting strategy for large loads and thinking about what opportunities exist to reframe our portfolio. We don't ever stop doing those activities, but trying to put a pinpoint date on things is just not something that we do. It tends to lead to expectations getting out in front of us. I think we prefer to do things, get them done, and then announce them. It's really hard for us to put, you know, we're going to do something by X quarter or say we're 65% complete or some number like that. We just don't do that.
I know that is probably not going to satiate your question, you know, scratch the itch, but that's where we are.
Okay. Secondly, PJM is getting ready for another round of reforms, or seemingly, of the capacity market. You made just basically flatlined capacity price expectations. Is it just a simplifying assumption? Is it just because you think that over time you're contracting more and more of your capacity and thus you become less dependent on the outcomes of these auctions? Any sort of outlook on PJM capacity prices?
I think as Chris Maurice mentioned, from the projections, what we did is just roll the cap forward. There is obviously the cap in the next coming up auction, which will settle in December, mid-December, as Chris said. From there forward, it was quite frankly a simplifying assumption with a lot of unknowns that are out there. We're obviously engaged in what capacity reforms might look like. We don't know if the floor and cap are going to get rolled to the next few years or if it's going to come off. Obviously, if it comes off, you saw that the last clear would have been at $3.90 a megawatt day uncapped. There is a lot of uncertainty out there. For us, rather than put a forward expectation in there, we just rolled it forward.
Just like we used West, we used the visible marks of West Hub and Tedco M3 as representative of the market because they're not our views. I think Chris Maurice said we're basically positioning the portfolio because we don't agree with where energy and sparks are in the out years. From a capacity standpoint, there are just a lot of moving pieces. We just rolled the $3.30 forward.
Okay. Lastly, on the tax shield. Assuming that the Guernsey and Freedom acquisition closes this year, you could potentially shed the taxes for 2025, I'm assuming. As of now, how long of a tax shield do you have, meaning like, you show us obviously cash taxes for 2026 and 2027 as a percentage of EBITDA. Can you disclose, for example, what it is for 2028 as well? Again, address the other 2025 cash taxes.
Sure. Good morning, Angie. It's Terry. You're correct. If we close the transaction in 2025, there would obviously be a knock-on effect to reduce our cash taxes further for 2025. However, I will mention that even with the new tax reform, our 2025 taxes do get a benefit from larger interest deduction and then also bonus depreciation. There could still be some reduction to 2025. With respect to the tax shield, it goes through the end of the decade. Obviously, we would love to be in a position to where we utilize that tax shield sooner rather than later. Now, you should just consider it rolling through the balance of the decade.
Very good. Thank you.
Thank you.
Thanks, Angie.
Our next question will come from Jeremy Tonnett with JP Morgan. Your line is open.
Hi, good morning.
Morning, Jeremy.
Thanks for all the details today. Just wanted to come back to buybacks and see if you might be able to expand a little bit more, I guess, on how you think about the pace of buybacks here. Is it more opportunistic versus steady state, or kind of it depends on what's in front of you, whether there's acquisitions? Just wondering any more color you might be able to share.
Yeah. Hey, Jeremy. It's Terry. I think past history is sort of indicative of where we're going to move forward. We've done both sort of opportunistic share buybacks as well as some structured deals throughout, and we'll look to do that as we move forward. Obviously, you know, we want to close the Freedom and Guernsey transactions first and get to a good spot there. You should expect to see a combination of both of those things.
Got it. That's helpful. Thanks. I was just wondering, as it relates to PPAs at this point, if you could speak a bit more to the relative appetite for nuclear versus gas and how that might have developed over time.
Yeah, sure, Jeremy. Hi, it's Cole. Look, I think if you go back two years when we were doing our first deal with Amazon Web Services and we were engaging a number of counterparties, there was a large focus on getting sites that had speed-to-market advantage, had the ability to get to a gigawatt scale, but also had carbon-free elements. That was nuclear. That was two years ago. After we announced the first deal, the market changed and the appetite is still for all three of those, but I would say that the nuclear carbon-free element kind of slid down the list a little bit from a must-have to a really nice-to-have. Speed-to-market advantages and the gigawatt scale, and as I said in the prepared remarks, getting up to even two gigawatts or three gigawatts over time at a site is just much more important.
That's where we've been focusing our conversations around providing those kind of solutions. Obviously, we still have a couple hundred megawatts of capacity of length at Susquehanna that we can provide through a front-of-the-meter contract, but build additional volumes in across our portfolio. As I said, also seeing a lot of interest in high efficiency, lowest heat rate premium combined cycles. We already have one in Lower Mount Bethel, and we're adding a couple more here as we close the Freedom and Guernsey transactions later this year.
Got it. That's helpful. Just the last one, if I could, talking about the power markets and the future curve. Just wondering, you know, your crystal ball, how you see things unfolding here given the flatness out there. Do you expect a gradual just kind of increase over time or catalysts to kind of lift it, or how do you think the market evolves at this point?
Yeah, I'll tell you, haven't dusted off the crystal ball in a while, but I'll give it a shot here. As I said in the script, I think the current forwards not being reflective of this tightening supply demand that we're seeing will get resolved through time. I think delaying that was the slow start to summer that we had. There was a lack of volatility, a lack of sort of intraday pops, which just limits some of that future shape that starts to play out. We haven't seen it. Is it gradual? Will it snap? I think cash events will be the driver of when we start to see some curve appreciation. We're getting into some sort of gas technical levels and seasonality on power curves, which will start to take shape. We've been on a small consecutive run of some up sessions here consecutively.
Certainly not claiming bottoms are in, but as price formation takes shape, there's a lot of constructive tailwinds that should keep that momentum moving forward. I think, Jeremy, I would just add on to Chris, which is when we put projections out there, we're always looking to use the most liquid and visible marks that we can so that it's a mark-to-market exercise. We did this as of, as I mentioned, the 31st of July. That's where these marks come from that underscore all these projections. That doesn't necessarily mean that we agree with those. It's just a standard way of providing outlooks on a go-forward basis. I think, as Chris mentioned, putting aside any gas move, which gas always moves fixed price power basically based off the market heat rate. Market heat rates have been higher in the past than they are today.
We see increasing demand, and it just, quite frankly, doesn't jive with where heat rates are out on the curve. Now, as Chris mentioned and as I mentioned, there's a recency bias always in the power markets, which is whatever just happened tends to either drag up or drag down the future. It goes both ways. It just happened that this summer was a bit soft. Not to belabor, but I think anecdotally that some other factors weighing in on that. There's the natural sort of market participant further out the curve. It's sellers, it's generators. That's going to sort of keep some ceiling on pricing. It's low liquidity. There's not a lot of trading or focus on, call it 28 forward time periods. Expecting or hoping that price formation would take shape in some of the outer years sometimes takes some near-term cash performance to drive that.
Just to piggyback on that, if we see where things are going, this is why we talk about large load contracting, not just data centers, but if you go back a few, I guess, more than a decade now, everybody's been leaning on the spot market. That's why you're using these types of marks. If you look at where things are going, when people are starting to sign 10- and 20-year contracts, we're going to head back to a world where CNI starts to sign longer-dated contracts rather than leaning on the spot market, which will then start to have price formation in the out years that won't be just sellers, as Chris Maurice mentioned, because that's really what it is. People that are out there hedging or a few speculators, right?
I think that, and it's not a very deep market once you get out to 2027, 2028, to be fair, at least on the power side. Gas is in a different position. We just don't see it being consistent. You saw the capacity market response, right? We just haven't seen the energy respond at the same time. As Chris Maurice mentioned, we're seeing really strong loads. We didn't see price formation, but really strong loads over the course of June and even the summer, even though there wasn't a lot of volatility.
Got it. That's helpful. Thank you very much.
Our next question will come from Nicholas Campanella with Barclays, and your line is now open.
Hey, good morning. Thanks for all the information today. Morning. Just going through the upside drivers, you kind of list out the one to two gigawatts of accretive M&A. Can you just kind of talk about, is this in jurisdictions in which you operate in? Are you looking externally outside of jurisdictions you operate in? Assuming that you do something in line with the Caithness deal, how quickly does the balance sheet reload after that?
I'll give my joke that I always give. I don't, you know, Terry has to worry about the balance sheet. I'm kidding, Terry. He's in here wincing that I start to say that. Look, we've already got the balance sheet being reloaded next year in the projections. Terry's been pretty quick about this, just to be fair, that we're looking to pay down debt. I think we've said something in the $300 million type of range that would get us back to the net debt less than 3.5 times next year. That clearly is our target, but the balance sheet reloads there. Obviously, it takes time to get to doing a deal and then closing a deal. We think that there's ample opportunity for it to reload or alternative ways. I mentioned perhaps we, you know, toggle and there's, you know, equity in order to maintain the balance sheet.
You can put all those things into the mix. To answer your first question, we, you know, look, we're always looking at opportunity. They don't have to necessarily be in our backyard, to use that expression. We do think that there are plenty of opportunities in our backyard to expand the portfolio and think about how do we add to, as Cole mentioned, this long-term contracting capability, low heat rate machines, whether they be Fs or H-type machines, but low heat rate because that tends to drive the lowest carbon production you can per megawatt hour of carbon producing generation. Lots of opportunities out there. We think there's room for further consolidation and ability to do accretive M&A that we can reload the balance sheet. It's probably in that, you know, that's why we size it as 1 to 2 gigawatts.
Yeah, Nick, to add to Matt's comments, when you think about the Talen flywheel, a core component to that is always getting the balance sheet back to an area where we could utilize it as a strategic asset, right? We've got in our projections a modest amount of deleveraging, but that then puts us in a spot where we can be opportunistic, we can engage in the market, and we can move forward with speed. You should just expect to continue to see that from Talen as we move forward, as we think about having that dry powder to use the balance sheet.
Okay. No, that's great. Just, you know, maybe quickly, just on your ability to add new megawatts at existing plants and any advantages or, you know, new lessons learned about, you know, where your existing gas assets are and the competitive nature of the sites from like a land availability or gas supply cost perspective?
Sure, Nick, as Cole, I'll take that one or at least start with that. Obviously, when we announced the upsizing with Amazon Web Services in June, we also talked about looking into upgrades at our facilities, especially Susquehanna. We've said it before, there's probably some minor ability to add megawatts to the existing portfolio. A lot of our assets, including Susquehanna, went through the major upgrades, you know, EPU upgrades, et cetera, a decade plus ago. I think from an existing fleet, the add-ons are modest. I do think we do have advantaged sites that over time, under the right construct, and by contract, I mean off-take agreements, we do have the ability to add new. We have pipelines that go right to all of our plants.
The Montour facility, we added a, we converted that to natural gas in 2023 and built a 17-mile lateral that we own that connects to the interstate. Same thing with Bruner, a little bit lower of a smaller of a distance. We converted that back in the late 2010s. We think that ability with excess gas capacity on those lines positions us well when the right time is to new build. That doesn't mean we're doing new build tomorrow. Obviously, as we've talked about, that's something we evaluate over time here.
All right, great. Thanks so much.
Thanks, Nick.
Our next question will come from Michael Sullivan with Wolfe. Your line is open.
Hey, good morning.
Morning, Michael.
Hey, Matt. Wanted to ask another one back to the large load proposal in PJM for future auctions. Just get a little more of your thoughts there in terms of where do you think this ends up going? What would be your suggested solution? How is the uncertainty that that has injected impacted any of your conversations, both in terms of future data center contracts and/or M&A?
That was a, you kind of fanned it out there at the end on the question, Michael, but look, I think when the contracted capacity back load proposal came out from PJM, there was a lot of noise around it. I think everybody's filed comments on it now. I think there's a general consensus that it's a little bit overreaching. What people are concerned about is how do you make sure that these load forecasts that are going to clear auctions are real, right? That was the real gist of that. There's a couple of ways to think about that, which is the first is we got to get the load forecasting, which is put onto the EDCs, needs to be somehow validated or made sure that there's some teeth to it, if you will.
If you look at, and Cole can expand upon this, there are teeth in what we're doing. We have a contract for 1920, right? We have, and Amazon has a contract with PPL to build transmission and to pay for that transmission, et cetera. There is the AEP tariff that goes out that's like 80% of whatever incremental costs need to be paid for, so you can't just walk away. All of those concepts are the same as saying capacity back load, if you will. That concept, there needs to be teeth in this because just putting down options, if people are just putting down options, it's going to stretch the grid further than real projects and potentially put real projects at risk with speculative projects. We think that speculative projects shouldn't be in the queue, or if they're in the queue, they should get a different treatment.
What is the solution to that? It's get the load forecasts right, and then let's figure out how to make that load forecast have some, make it tangible, real. How do we define that that should go into the capacity auction on the demand side for future years? I think that's where most of the comments that were filed were focused on. How do we get that right? We can go back and talk about the proposal and say it's far-reaching, it's discriminatory, it's all these different things. I think that's water under the bridge because everybody's stepped past that and said, how do we solve the real issue here? It's just that people look at it. Chris showed all this, the data center demand that either comes out from the Department of Energy or PJM itself regionally, all those different pieces.
How much of that's actually going to show up and where is it going to show up? That is the issue that's being tackled right now. We think that there just needs to be so-called teeth in those load forecasts.
Okay. Appreciate that, sir. I know that was a lot there. In terms of the Amazon Web Services deal ramp, definitely appreciate kind of the minimum commitment assumption in there. In terms of us being able to see evidence of acceleration and quicker ramp, what should we be looking for on that front?
Hey, Michael. It's Cole. Look, I think the evidence would ultimately be in the physical build-out, both at the Susquehanna campus and other campuses across Pennsylvania. Obviously, there's at least one other that was announced back in June outside of the Susquehanna campus. That's ultimately where you get to it. Obviously, construction is long lead, right? While we're a year plus into the first agreement, and there's certainly a lot of activity, not just with the shell that we built, but multiple other buildings in process of being finalized and future site plans already building out, as evidenced just by public view from the road and from parking lots that are not on their campus, suggests a significant build-out's coming. Until that build-out materializes, that's something that we're going to continue to model in the minimum commits until that power draw is closer in time.
To Cole's point, it's both the activity that we see on our site, Michael, that would lead to that, as well as, and Cole mentioned this, when we expanded the contract with front-of-the-meter, we also have the ability, or Amazon has the ability, to draw that power and direct it to other places inside of Pennsylvania. It's a combination of those things. You know, it's a great question. I don't know that there's an exact road sign that we're going to be able to point to for you on that, but we'll take that into consideration.
Okay, great. Thank you.
Our next question is going to come from Julian Dumlin-Smith with Jefferies. Your line is open.
Good morning. Thank you very much for the time. I appreciate it, guys. Good morning, Matt. Just on the free cash conversion, I don't want to needle you too much here, but is there any potential to see that improve? I know you have an analyst day following an acquisition that didn't change, but is there any ability to improve that 55% here at all over time as you think about it, again, versus the initial planning assumptions?
Julian, as Matt mentioned earlier, we're always looking at ways to improve, and free cash flow conversion is one of them. We'll continue to look at ways, whether it's capital costs, how we look at capital spend, the debt structure, any number of ways. Always looking at ways. I think you take a look at some of the things that we've done around mitigating credit costs and mitigating collateral costs over the last couple of years. That's sort of in our DNA. You should expect to continue to see us do that across the board.
Hey, Julian, I'd say 55%, 60% that we grow into is pretty darn good, quite frankly. This is FCF AG, not BG, so it's all in. The one thing that I would say further is as we grow into that last question about the megawatts, we have 480 in here. As we get to 1,920, that will obviously, because of the higher cash flows associated with that, convert to more cash for every megawatt. If we do additional contracting opportunities, that will convert to more cash per megawatt. As we change the fleet profile, if we're able to replicate what we did at Freedom and Guernsey, which have a high free cash flow conversion given the nature of the baseload asset, as Cole said, it's like adding another nuclear plant to us. Because of the low heat rate, that has a high free cash flow conversion from EBITDA.
All the activities that we're doing, including just growing into the existing contract, I think will increase that. I still think 60% is pretty good. I don't know if you had a different target in mind.
Only as much as you're hosting an analyst day subsequent to the announcement.
I'll just ask you for more. Don't get me wrong.
Julian, you know that.
Look, actually, speaking of asking for more, how about this Amazon ramp? I mean, you guys specifically call it out in your slides now about 2028 and suggesting that as an upside. I mean, you know, I imagine you wouldn't have put it in the slides if you didn't see it as like a real, you know, a conceivable outcome here. How do you think about that pull forward on timing for the Amazon contract, whether that's 2028 or 2027 or what have you?
Look, a couple of things on this. First of all, I think everyone out there knows how we do things as an entity, which is we're going to put the minimum commits out there. We're going to show what the potential upsides are. As far as do we think that it's credible that they could pull things forward? Yes. To the extent what the number is or the volume on that, you know, that's up to Amazon Web Services and how they build out the campus at Susquehanna and other campuses across Pennsylvania. But Cole, you?
Yeah. Look, I mean, as Matt said, we do not control the ramp, so we are not going to build in anything above the contractual obligations. As we said throughout, we have always believed that the minimum commitments were kind of an underwriting case. The PFIT, the outcome was an acceleration. That was part of the math and part of the allure of ultimately selling the campus to somebody who can build this campus as fast as possible, which allows us to sell more power. You can go back to a recent earnings call, the Amazon CEO talking about the biggest constraint for their Amazon Web Services growth is access to power. The faster they can build out at Susquehanna where they have the power, we think it is pretty constructive signposts that it could accelerate. Again, it is not an obligation, but it is the ability.
Again, it is not just at Susquehanna. That is what gets me really excited, the ability to flex this to other sites that can be built out in parallel. If there is a bottleneck on the Susquehanna site for phase two or phase three, just because of all the construction, they are obviously building at other sites, right? Susquehanna is not the only site Amazon is building across the country and certainly Pennsylvania. Hopefully that gives you some more color.
To be clear, there's not a bottleneck right now. He was saying hypothetically. Look, let's remember how we got here. I think this is a credit to Cole and the team that worked on restructuring this contract, the front-of-the-meter and even the behind-the-meter construct, which is we provided, it had to be a win-win for us and Amazon. We were the ones that were able to come up with a contract that said, look, you can take, there's going to be some intakes. That's the minimum commitments. You have the opportunity to flex the contract. That's because of, as Cole just mentioned, the ability to want to go faster if they can get more sites on and approve, build faster at our current site, et cetera.
It is that ability to work with customers and to deliver on customer needs and the ability to commercialize that, that I think led us to this point. Again, we forecast the mens and we look forward to exceeding those.
Yeah, absolutely. Just lastly, if I can quickly, on the cadence of deals, you wouldn't expect to do more M&A prior to closing the current acquisition here. I think the leverage commentary earlier probably suggests that as well. In terms of the cadence of, in fact, would you contract these first before doing M&A when you think about the next upside?
No, that's a good question. The question about timing with respect to would we do something while these deals are pending, we're always doing stuff, Julian. As far as would we contract these up before having to do the next deal, I think when we said when we did this deal, in a perfect world, what you would do is you would add an asset, you would contract the asset, then add an asset, then contract an asset. You can't do that because the world's not perfect. The world's lumpy. We're always looking at what are our contracting opportunities and what are our portfolio reshaping activities and balancing those. I think more importantly, look, when we did the Freedom and Guernsey acquisition, we underwrote that at Merchant, right, with the other side.
If there's an ability to go contract it or use it in the portfolio to contract. When we think about M&A activities, I think there was an earlier question which I joked with Terry. It's more about how do we manage the balance sheet and reload the balance sheet and do those types of things. Obviously, to your point, which you make a good point, which is if you can contract, you're creating more and more dry powder on the balance sheet. I think we've used the phrase we toggle between these different activities: balance sheet, capital allocation, acquisition. We're doing all things at all times. I don't know that there's a rule of thumb that we could tell you about contracting versus adding assets or even, quite frankly, selling assets like we've done with some of our small, you know, Camden and Dartmouth. In any event, I hope that helps.
Totally. Thanks for clarifying. I'll pass it here. Thank you, guys. Thanks, Julie.
The next question will come from Shar Pereza with Wells Fargo. Your line is open.
Hey, guys. Good morning.
Morning, Shar.
Good morning. Morning. Hey, Matt, can you just touch a little bit on some of the state-level kind of the resource adequacy bills that are out there, like in Maryland, New Jersey? You know, it's got an assembly bill. Pennsylvania's got a House and Assembly bill. Obviously, it could impact fundamentals and sentiment for the group. The understanding is we're kind of nowhere near new build economics and LODIS and going away, right? That's kind of the key things. Your wires peers don't really think in economic terms. I guess how is all this going to evolve as we're thinking about the bid ask there? Can you strike a middle ground with the wires companies? I guess how are sort of the negotiations going with stakeholders in those states?
Yeah, look, I think, you know, obviously Maryland has its process by which it's looking to secure LOD. We're providing the RMR there. That's what we're doing. We're looking at is there an ability to repower that site? The key to that site is getting gas there over time because of the pushback with respect to coal and to oil. We're doing our part with respect to the RMR. With respect to New Jersey, which is also short generation, we don't have sites there other than the Camden site, which is currently under a sales agreement and will be sold very shortly. We're more focused on Pennsylvania. I think the thing that's interesting about Pennsylvania, and we can get this for you, but there were some recent comments by PUC Chair DeFrank about the need for Pennsylvania basically to win.
I'm going to, this is paraphrasing, but to basically win the data center race and that we should use Pennsylvania electrons to power Pennsylvania data centers and get the economic growth there because we're a long generator. I think that that is very constructive from our perspective, and we've been part of that conversation. With respect to getting new build, I think every, you know, you said the wires aren't necessarily economic animals. I would think that they are to some extent because people have to worry about consumer rates at the residential level, and we're concerned about that too. We think that there are ways to solve that, but the market has worked and it is working and needs to continue to be allowed to work.
We think about when you look at what's going on with respect to the BRAs clearing at $330 and if they clear $330 this December and if there's caps off next year and they clear higher, that will incent some activity. There are already activities going on. I've always said that I think in the next five years, or we've always said, we think in the next five years, it's not an energy problem. We're running our units more and more. Montour is running more and more. Martin's Creek's running more and more. It's not an energy problem. It's a capacity problem. It's that 20 to 50 hours. There's different ways to solve that. We do think that demand response, not mandatory demand response, as was previously suggested, but a demand response element needs to be put in.
We also need to work and think about how do we put in longer-term contracts and think about that from a capacity market standpoint so that we can get some new megawatts built. People always go to the CCGT. I actually think that there's going to be much more constructive solutions that focus more on CTs and batteries, for example, in the near term over the next five years. CCGTs are going to be needed for the energy. We're focused on Pennsylvania. I think Pennsylvania's been constructive. I don't see re-reg happening in Pennsylvania. I don't see it happening in Ohio. We think in the areas that we're dealing with, we like where we are.
Got it. Perfect. Lastly, the 28 upside slide, I know you kind of highlighted that you weren't going to comment on the one gigawatt data center opportunity. Maybe is this sort of a wishlist item like the CCGT M&A opportunities, or is there some real traction here, especially as we're thinking about ESAs and LOIs? I guess we're in the process. Are you at there?
We're working on them every day. Cole, you want to tell them what we're doing? Thanks for not commenting on it.
If Darren was here, I'd ask him to elaborate on which activities we're going after. Look, there's a lot of different combined-cycle gas turbines and assets that are coming to the market, quite frankly, in our backyard, which we've expanded our backyard down into Ohio. As I mentioned, I don't know that there's been a, I mean, is there a day that I haven't asked you, what are you doing? Like, when are you going to get something done? Anyway. Might be a few out there.
A few days, maybe a few hours in between. We're always working on all those things. I think people are starting to rationalize that these things aren't as simple as you just take an off-the-shelf contract and sign it. We like that, quite frankly, because we have a competitive advantage. It's a little bit of a barrier to entry. We know how to warehouse the risk. We know how to meet customers' needs. I think that was part of what I was saying is that there's a lot of opportunity out there on the contracting front. I think there's an opportunity on the asset front because we have headroom, we have ability to grow, and we have the platform to attach things to or to add on, bolt on, if you will.
Got it. Fantastic. Thanks, guys. See you in a couple of weeks. Appreciate it.
All right, Shar.
Thanks, Shar.
Our next question will come from Emma Cutchi with Evercore. Your line is open.
Great. Good morning, guys. That wasn't the worst pronunciation of my last name, so that's good. I've been called a lot worse, especially by my wife. Just wanted to, forgive me if this is kind of a dense question, but when we think about kind of the acceleration of the Amazon Web Services deal, right, the potential acceleration, I'll say, of the Amazon Web Services deal and the flexibility across assets, when we think about that and then kind of a fixed amount of megawatts associated with them, what's to keep kind of like an Amazon Web Services from then? Would you be able to recontract kind of the balance if they flex it out to different sites and leverage different assets? Are you then able to recontract that at Susquehanna? I'm just trying to think through that and make sure I understand this fully.
Yeah. Just to be clear, the current PPA that's now nearly 2 gigawatts allows the flexibility to flex utilization of that PPA at other sites. Your question is, what happens if they get to the full, you know, 1,920 megawatts across multiple sites and they still have excess power needs at the Susquehanna site? Look, I mean, that's going to be something we discuss. How do we upsize and how do we accommodate? That's something that right now, the contract's limited to the 1,920. As you know, we've talked about, we have length across our portfolio. I don't want to get ahead of those conversations. They still need to build out 2 gigawatts before we get to that point.
Right. Yeah, good problem to have. Understood.
Perfect. If we think about, obviously we're not going to comment on specific M&A, how about the broader competition within the M&A market? We've seen some creative M&A type of avenues taken by some of the regulated players. When we think about these long-term, long-dated, fixed, regulated nature type of contracts, are you seeing more interest and more competition for assets from the broader utility community? Not necessarily in the utility community to date. There has been some speculation of utilities participating through a deregulated subsidiary. I do think that the market for assets has picked up. I think there was a general trend over the last two to three years in a lot of the private equity shops to basically sell or to move on. There were a lot of assets that have been under hold for a long period of time.
You saw several big transactions earlier this year with LS and Calpine, and then our transaction, which was very meaningful for us. I think you're starting to see people go and raise funds and get back into, I'll call it, traditional thermal generation. The move has been back in. I think that goes to the point of talking about that it's a pretty exciting time, whether it be in Chris Maurice's job, although sparks haven't responded yet, but in the M&A type realm, as well as what we're doing with the contracting strategy that Cole Muller's managing, there's just been a lot of activity. Terry.
Yeah, I just, I mean, Max hit on an important point. You've got a lot of private equity firms that obviously participate in this space. They have transactions where they're recycling capital, and you should expect that they'll re-engage in the market as that capital cycles through, just sort of a natural cycle for those guys. Obviously, the strategic transactions that you've seen, when you take a look at the strategic buyer universe, you've seen our peers like Constellation and Vistra do transactions as well. I think healthy amount of demand and competition out there, and obviously we've engaged in it as well. That's sort of the state of play, if you will, in the M&A space.
Perfect. Thanks. Just one last quick one, kind of cleanup question. Should we read the flow-through of the $329 cap, obviously within the collar through 2028, as kind of a conservative estimate? Is that the intention, that just given the tightness within the market, we'd expect that to be conservative?
I would read it as just a simplifying assumption, right? As we've heard on this call, there's obviously questions about, you know, not for this next coming auction, but the auction after that, you know, what happens. To sort of reiterate some of the comments that Mark McFarland mentioned earlier, I think one of the challenges with the capacity auction and one of the things that we hope to see is a continuation of the schedule because there's nothing more that hinders long-term capital investment than uncertainty. Keeping that schedule and keeping it moving forward is the key for us. It's just a simplifying assumption there.
Our next question is going to come from Ian Zaffino with Oppenheimer. Your line is open.
Hi, great. Thank you very much. I just wanted to ask on the Susquehanna and the outages. How are you thinking about next year as far as unit one? I guess that's above the typical cost that you kind of lay out. What's basically driving that? Do you think maybe unit one may eventually cost as much as the unit two outages costing this year? Thanks.
Yeah, Ian, in the appendix on slide, I think it's 23, we lay out some details around that for the outages coming spring on unit one.
We do expect the outage cost to sort of be in line. Two things. There's O&M cost and there's capital cost. We expect them to be in line to slightly under what we've experienced so far. I think the big gain that we'll get is the number of outage days that we'll see. That's sort of the expectations. Obviously, we've learned from the extended outage that we had earlier this year. Slide 23 lays those things out. We think that we'll have a lower lost opportunity or lost margin because the outage will be done quicker and then slightly lower cost at the end of the day.
Yeah, Ian, just to piggyback on Terry's there, the slide lays that out. I think what's important is that we're going to perform the same work that we performed on unit two this year. We've learned from it. We've shortened the outage schedule for next year. What I would tell you is that we return to what I'll call normal. There's no such thing as a normal, but a customary outage, a typical outage, as we represent on the appendix slide in 2027 and 2028 in the forecast. We think we'll get the same megawatt, you know, close to the same type of megawatt recovery. We're performing the same type of work at the site this outage in March, April, excuse me.
Okay, thanks. Maybe as another question here, you know, and I know this has kind of been asked in several different ways, I think. When you kind of lay out that there's one gig of upside on a PPA in that slide, how are you kind of arriving at that, right? There's the extra Susquehanna that potentially should be signed. It's not just because it's not going to roll in by 2028. Is that kind of how you think about it? You also have two other facilities out there that I think are also available for PPAs. Why just kind of pick one and not the other opportunities? How are you getting there? What's actually the confidence to get that? Any other kind of thoughts you could give us as far as also negotiations with gas costs versus the PPA price? Thanks.
Yeah, thanks, Ian. Look, I mean, I think as Max said in the prepared remarks, we just made a simplifying assumption of one gigawatt to illustrate the impact of that. Obviously, the first two deals we did, each were incrementally roughly a gigawatt, but that doesn't mean that has to be the size going forward. Second, I think just to be clear, if we're doing front-of-the-meter deals, we don't have to do them asset by asset, but we can kind of have a basket of assets backstopping a single volume for a deal. Whether that deal is 500 megawatts, 1,000, more than that, we can backstop that from a kind of a blend of our Susquehanna lengths that we still have, pick your next plant, and then your next plant to get to the total. That gives us a lot of flexibility.
Obviously, as we close on adding Freedom and Guernsey, that just gives us more flexibility, not just to backstop an existing deal that we've been working on since well before that announcement, by the way. This gives us additional dry powder in terms of megawatts to contract in the future. I would not look at that slide and say there's only one gigawatt or there's going to be one gigawatt. We have a fleet. We're looking, as we continue to say, we're looking to contract that fleet up to a certain %. It won't be 100%, but more than we're at today. You can see kind of the rough impacts of a gigawatt, hypothetically. I think it's very interesting. The reason why it's standardized as this hypothetical in one gigawatt for the data center is that if we did two, you could expect twice the free cash flow accretion.
It's just there so that you can do the math.
Yeah. In your other parts of your question, I'm not going to comment on where we're at in the negotiations. We obviously wouldn't be talking about things if we didn't think we were advancing. I said that in the prepared remarks, we're advancing. You had a question in terms of gas, and maybe I'll start and kick it over to Chris to talk a little bit about our platform to manage the gas piece. Obviously, that's an interesting variable or an added variable that we didn't necessarily have when we're talking nuclear power, or at least the volatility there. The short answer is, without getting into the details, it really just depends on who's taking that price risk, right? If we're not taking that price risk, then it's not really anything we need to manage.
If we're taking the full price risk, then it is something we're going to have to manage. Maybe there's something in between to share. I think each counterparty is going to have a different view on how certain they want their pricing and how much risk they want to take. We have a platform. Chris and his team can manage that. We think we are a natural counterparty to warehouse appropriate risk there and obviously get paid an appropriate risk premium to do that. Chris, you want to talk about?
Yeah, no, you hit it. I think, again, sourcing physical molecules to power plants is something we've demonstrated a capability, you know, for decades now. Flooring over a B of gas on normal peak demand dates for us, layering in a new data deal.
Again, there's some nuance and some complexity to the financial side of managing that price risk, but again, a construct we're very comfortable with.
Okay, thank you very much.
The next question comes from Rinny Singh with Bank of America. Your line is open.
Good morning, guys.
Morning, Rinny.
I guess just my question, my first one would just be about, you know, we've seen a lot of large load announcements in the vertically integrated market. It kind of based on tariffs, it seems like there's a cost advantage. At the same time, you know, long contracting times and then regulatory hurdles seem to be slowing the speed to market advantage in deregulated markets. I guess are you still seeing the same demand in that PJM market, or are you kind of seeing them move to vertically integrated? What advantages are still there for deregulated markets versus the vertically integrated, you know, if we have these regulatory hurdles still in the way?
Let me, I think that there have been things discussed as regulatory hurdles.
I think that if you look at Pennsylvania, Pennsylvania is open for business with respect to data centers, whether it be Governor Shapiro, PUC Chair DeFranc. There's a lot in the local support in and around Susquehanna. Cole can talk to this as well as other areas. I don't see there being necessarily regulatory hurdles. I think there is a concern about how does all this manifest itself and make sure that we protect consumers, etc. I think that speaking specifically to PJM and things of that nature, there's been noise around, for example, the capacity back load proposal, etc. I don't see there necessarily being regulatory hurdles. I don't think that anything's changed. I think that the, Cole, wouldn't you say that the activity level has remained basically the same?
Yeah, I would say, I mean, I wouldn't even say it's the same. I think it's elevated. Look, you, I won't pick on that specific utility, but utilities in our general area have been talking on earnings calls about continued interconnection requests going up at advanced stages. Folks can go to the PJM website, deregulated market, and look at certain utilities and what's been submitted from a public standpoint. We're talking substantial peak watts. The reality is when a data center or any load signs up in an energy supply agreement with the local utility, they don't announce that, right? I think there might be other reasons for a broader announcement within the vertically integrated space where you have the generation and the connection all in one.
In our space, the IPP space, when we have a big generation or energy and capacity agreement, obviously, we need to disclose that from a materiality standpoint. The hyperscalers, by and large, aren't going to sit there and say, "Hey, we just signed up another three gigs to connect to region X, utility X, Y, Z." We're seeing plenty of signs that certainly in our local area the speed to market advantage is real and it's continuing to accelerate.
I also think Pennsylvania, as DeFrank made this comment, is situated right there between basically data center alley in Northern Virginia and a lot of load. It has to be one of the areas that PJM has to be one of the areas where data centers go. The natural selection there is Pennsylvania, and Pennsylvania has made a commitment to it. I do think you see announcements in the vertically integrated, but I think we offer something different, which is what we've been focused on, which is when you put a data center down in Pennsylvania and do the type of contract that we did, what are we offering? We're offering basically going out and fixing price and delivering price over a period of time. That doesn't happen necessarily.
You know what your prices are when you go in and you pay for your transmission and distribution, but energy and capacity can change in rate makings with a new rate making, a new case, or even a rider that goes in. When data centers are looking at this, they're also looking at deregulated markets because then you get a fixed cost structure for a decade or two, like in our contract with Amazon Web Services.
Okay, that makes sense. I guess the certainty and location are kind of big drivers. I think you mentioned on the earnings call about the diverge between the transmission portion of bills and the generation portion of bills. Could you just touch on that a little and what we're seeing there that's kind of really driving the affordability crisis?
Yeah, look, I think I'm going to try to pan back out a little bit on this and just talk about this, which is like it's obvious that when you get increased load, you need from data centers or large loads, you need to balance. I think Pennsylvania is doing a great job of this so far. I think that our work along with PPL is doing a great job with this, which is you need to balance adding the economic development, the increased load with making sure that consumers don't bear the brunt of that. I think that's where the big concern is. If you look over time, you know, rates have gone up, they've gone down, etc. People are looking at the forecast going forward and say, how do we protect the consumers, but how do we get the economic development and how do we serve large load?
If we can do that collectively together, that's when everybody wins.
Okay, that all makes sense. Thank you so much, guys.
Thanks, Rinny.
Our last question comes from James Wetherbee with Milius Research. Your line is open.
Thanks. Good morning, guys.
Morning, James.
Hey, morning, Mark. Quick question for me. I know we're way past an hour. I think we're at an hour and a half now, so I'll be quick. I think it's pretty straightforward. As you think about duration, and especially the lack of duration in kind of hedging markets versus the duration you can get from a PPA with a data center or a longer-term agreement with CNI, what's more, I guess, desirable for you? If CNI is, let's say, shorter, maybe it's 5, 10 years, then you reprice into a much better market, or is taking a 20-year data center contract more agreeable?
I think that's a great question. I think that's, as I said earlier, where things are going, is going back to sort of long, more longer-dated contracts in the CNI. It's probably a little bit maybe out of sequence if you were going to do things logically. By doing the, call it, 20-year contract with Amazon Web Services, it puts a focus on it.
Right.
You know, I do think where things are going back to is that prices, as prices rise in the energy and capacity, what's fallen away is the actual buyers of those three to five-year contracts. We have to have the buyers of the three to five-year contracts come back. We think that's going to happen, and we're set up to do that, which is the old origination function. It's originating and selling. We have a retail license in Pennsylvania, and that's how we're executing under this AWS contract. We could easily do three, five, seven, ten-year CNI deals. Your question then unpacks to what are the most advantageous of doing those. I think it just depends on the willing buyer, willing seller, the price, the risk that we're willing to warehouse for somebody in order to either give a fixed price or share risk as a flow-through.
We're pretty excited about the opportunity. Obviously, we've been focused on doing gigs at a time, as Cole's done, and what the next data center deal might look like. At the same time, we're thinking about how do we serve the longer-dated three to five-year contracts with respect to CNI.
Yeah, the mix of how and where we'll be able to hedge in the future will continue to evolve. Less screen trading, fixed price hedging, more contracted PPA types is where it's turning.
Great. Makes sense to me. Thanks, guys.
This does conclude the Q&A session. I will now turn the call back over to Mark for closing remarks.
Great. Thank you, Michele. Appreciate everybody's time today and your interest in Talen, and hope you all have a great day. Thank you.
This does conclude today's conference call. Thank you for your participation. You may now disconnect.