BP p.l.c. (BP) Q3 2025 Earnings Report, Transcript and Summary

Welcome, everyone, to BP's Third Quarter 2025 Results Call, which today we're hosting in Abu Dhabi. We'll be focusing today's call on the third quarter results and the contents of the video that I hope many of you will have seen by now. But before we move to Q&A, let me firsthand over to Murray for a few brief opening remarks. Murray?

Thanks, Craig, and thanks, everyone, for joining the call today. We're now 3 quarters into our 12-quarter plan and have delivered another strong quarter of operational performance and strategic progress. Earnings and cash flow generation was good with underlying pretax earnings of $5.3 billion and underlying net income of $2.2 billion. And with $7.8 billion of operating cash flow delivered this quarter, we are making good progress in delivering on our growth target for adjusted free cash flow growth of 20% CAGR over '25 to '27. Our operations teams are doing a great job in running the assets well with upstream production increasing by around 3% quarter-on-quarter, supported by upstream plant reliability at around 97%, leading to upgraded full-year underlying production guidance and refining availability also close to 97%, the best quarter in 20 years for the current portfolio. Looking ahead, we're also making strong strategic progress. We've started up 6 new oil and gas major projects in 2025, 4 of which were ahead of schedule. And we've had 12 exploration discoveries so far this year, including Bumerangue in Brazil, where the latest analysis and results gives us even further confidence. This performance is also showing up in the downstream as well. Underlying earnings in the first 9 months were around 40% higher than the same period in 2024. In customers, we delivered our highest 3Q on record and refining captured a better margin environment. We're making good progress on derisking our $20 billion divestment proceeds target, today upgrading our proceeds guidance underpinned by proceeds completed and announced this year that are expected to be around $5 billion. We're staying disciplined with our capital investment with organic CapEx on track to be below $14 billion. And we remain confident in the momentum we are building in support of the delivery of the cost and net debt targets we have laid out. In summary, while there remains a lot of volatility, we are staying focused on what we control, underpinned by a laser-like focus on performance across the company. We have world-class assets and capability with operations delivering strongly. We have a deep resource base and are building high-quality options for growth in the future, the focus of our ongoing portfolio review. We're continuing our momentum to drive -- in our drive to reduce costs. we're making good progress in growing cash flow and returns and our plans to strengthen the balance sheet. And of course, we have more to do. All of this is in service of growing shareholder value and returns. With that, I'll hand over to Craig to take us through the Q&A.

[Operator Instructions] So I think we'll start there with taking the first question from Al Syme at Citi.

Murray, I got a question on Bumerangue. I'm really intrigued by the decision rather to publish the map on Slide 9. Clearly, there's a lot of market interest in the discovery, but at the same time, it's quite early days to be publishing a map. Can you talk about the confidence in that geological map based on the data you've got? And I'm also intrigued to know whether that sort of image looks any different to the predrill assessment that you had in the field.

Yes. Great, Al. Thanks very much. Yes, we're feeling pretty good about Bumerangue right now. As we disclosed recently, 1,000-meter column, 100 meters at the bottom of oil and 900 meters of rich gas condensate. We've evaluated about 2/3 of the samples in the labs, and we continue to evaluate them. The map we've produced is off the predrill seismic. There's a lot of technology that's changed over time, and the pre-drill and post-drill are pretty close to each other. The guys were able to image the top and bottom of the reservoir within a couple of feet based on the quality of the seismic we had. So we're feeling pretty good about it. It's a pretty good aerial view of it, 30 -- at least 300 square kilometers, at least 1,000 meters of column height, and we continue with the lab sampling. We will update the market in due course once we understand the gas oil ratios fully and once we understand the volumes in place. And we've secured a rig to drill the next appraisal well and do a flow test on it as well, which we expect to happen once the equipment is available near the end of next year. So good news on Bumerangue. Thanks for the question.

We'll take the next question from Alejandro at Santander.

The question is about Castrol, the process, the strategic review of this asset. If you can give us some color about how the process is going.

Great. Thanks, Alejandro. I'll take that one again. First, just a small note of congratulations to Emma and Michelle, who run the business. It's 9 quarters in a row of increase in earnings, very strong performance out of that business, and it's going very well. It's a commercial process, so I won't talk much about it other than to say there's strong interest. We are moving at pace, and we'll update you in due course. You'll remember that any proceeds that come from the strategic review will be dedicated to the balance sheet. But strong interest. We're moving at pace, and we'll update the market when we have something to say to the market directly about it.

We're going to take the next question from Irene Himona at Bernstein.

Congratulations on the numbers. Murray, can you give us an indication of an approximate timing for making concrete announcements to the market on the further portfolio simplification and restructuring, which you referred to in your comments, please?

Great. Thanks, Irene. Thanks for your kind words. On the portfolio review, Albert New Chair is on board now. We're starting to work with him on thinking about the portfolio. Of course, this comes about because we've had such tremendous success inside exploration. When we set out our plans in February, we didn't imagine that we'd have 12 exploration discoveries in a year. We certainly didn't imagine we'd have a discovery like Brazil as well. So that's all good news. We, of course, as a corporation, are very focused on making sure that we drive for value and returns and allocating capital to the highest quality opportunities. And that's what we're commencing now. We plan to update the market as we go along. So if you think about what happened in the third quarter, you saw that we made a sanction decision on Tiber in the Gulf of America, which we're very happy with. You saw that we decided to divest the Culzean field in the North Sea. We feel that it would have more value in other people's hands. And you saw that we stopped the Rotterdam biofuels refinery. It just didn't -- it didn't compete on a returns basis in our portfolio. So we'll update as we go along, Irene, and you should expect us just to update us as we go along through time and the decisions that we make. Thanks for your question, Irene.

We're going to take next question from Lydia Rainforth at Barclays.

So can I just come back to Bumerangue, if I could. Just on -- what I'm getting back is a lot of, well, it might not be 300 square kilometers, you may only be able to access half of that, the CO2 content. And like every is trying to talk it down. So just to take a step back and just on your earlier answer, Murray, the idea of the commerciality, that was really the main message that you wanted to share with us some update last week. And then secondly, just a very different topic on AI and the cost base. We've seen lots of examples here in APAC around just agentic AI, what we've seen there. Can you just talk us through what -- how you think the deployment is going within BP? And you talked about wanting to reduce the complexity of BP, improve the simplicity. So can you just walk us through where you think you are on that journey?

Yes. Thanks, Lydia. Just on Bumerangue, I think the principal thing to focus on is that there's an awful lot of oil and condensate in the column. It's a large column. We've updated it from 500 meters to 1,000 meters based on the logs and the strong response we've got inside the logs and the samples. We do have the 100 meters of oil. We do have the 900 meters of rich gas condensate. That makes the CO2 manageable, although you might need a little bit more money for metallurgy. Obviously, you're going to get an awful lot better flow with CO2 in an oil condensate column. So we feel comfortable. We continue to think of it as the largest discovery in 25 years. We've obviously secured a rig to go appraise it and test it, and we feel that we're increasing the quality of this with the results that we're seeing out of the lab moving forward. And we'll, of course, update you on gas oil ratios and volumes when we're ready with it, Lydia. I think on AI, I do think we're making decent progress. A couple of quarters ago, I had Emeka talking to the sell side about what we're doing in AI. And all of you know that we've partnered with Palantir more than a decade ago inside the upstream to really get going on structuring our data and experimenting first with linear programming and then moving into AI more recently as that technology has emerged. I think the first thing to say is we feel well progressed on the data foundations, which is critical to making AI work. We've said that we'll have a unified data platform, not just across the upstream, but across the downstream, across trading, across finance, where working with Palantir and Databricks, we'll have an entire unified data structure sometime around the middle of next year that then allows us to use AI and the LLMs against all the data we have. That's quite exciting that we'll be in that position. It's all cloud-based, so accessible everywhere. And I think that will make us distinctive for having that type of data structure on topology. The actual examples of AI that we've got going around the company, I feel good about as well. Last quarter, I talked to you about kit detection, where the teams have worked with the LLMs to be able to predict kits ahead of meeting them while drilling in wells like far south in the Gulf of America, and we're at about 98% detection on kits. As well, you saw in these results, production is high. We've upgraded our production guidance for the year. Why? Because we're at nearly 97% availability in the upstream. That's the AI helping us predict faults before they occur, repair them before they occur along with all the investment in the hydrocarbon kit we've done. That 97% is a record across since merger time. So outstanding result. And we're also seeing that inside the wells. Wells are failing at a far less frequency than they were as the AI helps us manage pressure depletion inside our well stock. Another great example is well planning. The AI is enabling us to knock down well planning by 90% as it catalogs all the data, provide suggestions to the experts, and that's significantly increasing the speed with which we can plan wells. not only safely, but more efficiently. And then it's not just contained to the upstream. We're working very hard with Palantir and Databricks to work our way through refining in a few of our refineries. And in the customers business, there's an interesting example of an AI agent that's helping us in our service stations in Germany. We've trialed it with 20 service stations in Germany. It's been designed to help us manage our stock levels there to make sure that food isn't wasted, that we follow customer preferences for what they like to purchase and what they don't like to purchase. And that after 3 months in those 20 locations, they've knocked down waste by 45%. So we can see tremendous examples of improved uptime, improved performance, better capital efficiency through AI. And we're very excited about the opportunity this has as our data foundations get firmly in place. Thanks for the question, Lydia.

We're going to move to the U.S. to take the next question from Doug Leggate at Wolfe.

Murray, I guess I've got a couple of parts to this related to your production guidance long term. You're already over 2.3. BPX is knocking it out of the park, frankly, versus its more than 600 end-of-decade kind of guidance. And now you've got multiple discoveries and potentially an early production system from Bumerangue. So my question is, how do you see the risk to your production guidance? And maybe a kind of part B to that, would Bumerangue early production system be included in the CapEx guidance that you've given us over the current plan as well?

Yes. Thanks, Doug. I'll hesitate to give a ton of guidance forward. We've set out our plans to 2027 as principal guidance, and then we gave an indicator of volumes at the end of the decade as well. I think it's probably premature until we work our way through more of the portfolio review to understand where we'll be headed. We do have choices on short-term versus long-term. So of course, we can pivot more capital into BPX and drive-up production near term or we can pivot more to things like the Paleogene and Brazil to drive longer-term resource production. I think if I step back from it, I think the thing I'd say is that I feel we now have the potential to grow long-term organic oil volumes for long duration. And I'm not sure I've been able to say that over the past 25 years with BP that we've been in a resource position like that. It's a nice problem to have. And what we're tightly, tightly focused on is staying within our capital frame and deciding what the right thing is to do to grow shareholder returns and value on behalf of the shareholders. So I think where I'd wrap that question is I'm very pleased to have been able to improve the guidance for 2025 after only 3 quarters, tremendous performance from the teams, as I said earlier. We'll update you on 2026 in February with what our viewpoint is of production then. And I would say we have more potential to grow, especially in oil now. And I feel we're in a better place than we've been in my career with BP, which is a nice thing to have. I hope that helps.

We'll take the next question from Lucas Herrmann at BNP.

A couple of -- well, a couple, if I might. Just going back or staying with BPX, Murray. Just trying to understand the CapEx profile. I mean, I appreciate the growth is very good. But the rig count kind of held. And I guess my understanding was always that as you came to complete on the processing facilities, bingo, so on and so forth, that we see a step down in spend, which doesn't really seem as though it's happening. So some explanation as to the developments there. And then if I can, just a simple one for Kate on the pension fund, if that's ever simple. Can you just talk around the buy-in that you've arranged with Legal & General and why you -- why sort of stopped where it has at this point? Should we be expecting you to sell down or to allow Legal & General to buy in a greater proportion of the U.K. fund into the future?

Thanks, Lucas. I'll let Kate talk, and then I'll come back on BPX.

Yes. Lucas, thanks for the question. So yes, I mean, it's been a conversation inside the Pension Trustee Board for a while in terms of derisking. You can see a number of other companies have stepped into this in similar ways, some to a bigger degree than we have. I think the transaction that's been executed is a good one to date. But of course, it's not our decision as a sponsor as a company. It rests fully with the Pension Trustee Board. And I think they will continue to evaluate how they feel with regard to further derisking as we go through the coming months. But I have nothing further to be able to say in terms of guidance on that at the moment, Lucas.

Great. Thanks, Kate. On BPX, Lucas, the way that we think about BPX is about $2.5 billion a year into it. Of course, we have the opportunity to flex that up and down. So this year, we'll spend around $2.5 billion in BPX. And as guidance, I think we guided around $2.5 billion through the next couple of years as well as we talked about the shape out to over 650 kbd in 2030. I think a few things I'd say about BPX. The productivity improvement we've seen from the team is very high. They've had 30% productivity improvement in completions and 15% in drilling over the past 12 months. They're now at top quartile in each of the basins we operate from a drilling days per 10,000 and they're at top quartile on NPV per dollar spent, which are fantastic metrics to continue to push and congratulations to the team on doing that. Another little advertisement for them would be they've drilled the best well in the Haynesville now ever, a 4-mile lateral completed and now producing 80 million standard cubic feet a day, which is a record for the Haynesville. So congrats to the team for doing that. As far as where do we go from here with BPX, we see continuous drilling inside the oil windows. You'll notice quite a large liquid growth 2Q on 2Q, '24 -- that's as we fill up the Permian, and we're doing an awful lot in the Eagle Ford as well. There's strong growth in the Eagle Ford from the infill spacing, the downspacing I've talked about before and from the refracs I've talked about before. So very strong liquids growth across that business. And the natural gas, the drilling and natural gas, we're running 8 rigs across. We'll have a conversation as we head into 2026 about do we keep running at 8 rigs? Do we move it up to 9 rigs inside the gas window? But the productivity improvements are so strong that they've actually drilled 13 wells basically for free this year relative to what our plans were. So I think count on 2.5 until we give you additional guidance. Obviously, more drilling than infrastructure as we finished off the infrastructure -- the major infrastructure program, as you mentioned. And we see the chance for a strong growth moving in BPX moving forward, and we're happy to support the U.S. in growing production. Thanks, Lucas.

We're going to take the next question from Chris Kuplent at Bank of America.

Trying to stick to the one-question rule, but a wider one. beyond Castrol, Murray, could you maybe let us know where you're at on Gelsenkirchen, on Lightsource? And if I may just ask, you've done the TANAP stake disposal now in BPX. How many of those midstream opportunities do you still see when you look across your portfolio for potentially more noncontrolling interest stakes?

Yes. Great, Chris. Thanks very much. We laid out a program of $20 billion. Pleased to report that we've announced $5 billion now. I think $1.7 billion of proceeds in the door case and obviously, another $3.5 billion to come in the door to help the balance sheet as we complete these transactions as we get to completion and approval. I think what I would say is there's a strong interest in Castrol, and we continue to move forward with that. We'll update you. Same for Gelsenkirchen, strong interest, and we'll update you. And on Lightsource, we started strategic conversations with counterparts, and we're at an earlier stage on that than we are on both Gelsenkirchen and Castrol. So you should expect something that takes a bit longer to disclose on Lightsource. But we're making strong progress on all 3 of those things. And we'll update you as the commercial processes, I don't want to say much more than that. We'll update you when we have news to tell you. I think on the infrastructure stuff, Kate, why don't you take that, please?

Yes, Chris. So as we look out in terms of the delivery of the rest of the $20 billion program, we don't see any other significant infrastructure deals in the pipeline. So in terms of how you should think about NCI, the way I would suggest you hold it is it's not going to increase beyond this. And actually, next year, as we redeem the hybrid that we prefinanced, there's about $1.4 billion left of the 2026 maturity that we prefinanced, if you remember, then that will start to bring NCI down. And then should we choose to take advantage of the 25% of the hybrid stack that we could taper under the S&P rules, should we choose to step into that space, then you'd see NCI reduce further. So the way I would suggest you hold it is it's where it is and from here, it will go down.

We're going to go back to the U.S. taking the next question from Ryan Todd at Piper Sandler.

Bumerangue is rightfully getting the bulk of the attention right now, but you've had quite a bit of success across the drill bit across the portfolio. Can you talk about what are some of the other discoveries or opportunities this year that have been particularly exciting and maybe in particular, in Namibia, what you've seen so far, how it's comparing to expectations and the timeline of next steps?

Yes. Thanks, Ryan. Let's see. I think maybe -- so we're -- I think we're 12 out of 14 right now, if my math is right, on discoveries. So congrats to the explorers for such a great year. It's probably the best year in our history. Particularly interesting has been the convergence of seismic technology with big -- with new chips from companies like NVIDIA and the emergence of AI. It's allowing us in places like Egypt, Trinidad, Brazil to see below salt much better than we ever have. And I can remember looking at the seismic on Egypt where you could actually see the channels. So there's -- we're seeing a change in technology that has helped exploration this year. I don't want to say it will necessarily do that next year as well, but that's been part of the story of the excellent exploration we've had. I think Trinidad offers up 2 good discoveries for development. Egypt offers up 2 good discoveries for development. Brazil, we've talked about. And if I then move to Namibia, again, we're very excited. That's been done through Azule, our joint venture with Eni. We've had effectively 3 discoveries. The third one, Volans, came this -- in the third quarter. We've got a nice reservoir in Capricornus is the second one, 38 meters, a very high Darcy rock, good oil properties. And then we have Volans discovery, 28 meters if memory serves, rich gas condensate, only 14 kilometers away from Capricornus. So Namibia is looking like a very good block. We continue to test the samples in the lab through the operator of the exploration phase Rhino, and we're quite optimistic about it. I think the Namibian Energy Minister called it the best block in the nation. So we're really pleased with that and looking forward to further appraisal and an update from the operator, Rhino, in due course about how we take the development of this block moving forward. But thanks for recognizing it. A very good year for exploration, and we're proud of the teams for what they've delivered. Thanks, Ryan.

We'll stay in the U.S. and take the next question from Paul Cheng at Scotia.

Murray, I want to go back into exploration. You guys definitely have a very good year. In addition to the -- maybe that combining AI and seismic to allow you to be able to see through the rock better. Is there any processes or the personnel changes that can lead to this great success? And how repeatable are they? From that standpoint, going forward, if you do believe that you have better success rate, should you deploy more capital into the exploration going forward to be able to use it as maybe a larger source of replacing your resource going forward?

Thanks, Paul. I guess there are a few things happening, first of all, inside exploration. We have a great experienced team who have been high graded over time, and they've built on the track record of their predecessors and built up a very good base of knowledge around the world. So we have great people with great deep knowledge, I would say, of the basins in which we operate. I think the second thing is technology is changing. The NVIDIA chips that we're now using inside our supercomputing are just incredibly fast and allow incredible iterations of theories. I'm kind of dumbing it down, all the geologists on the call, please forgive me for dumbing this down. But it enables much faster interpretation ideas, thinking about how one can think about the subsurface. And that, of course, is converged with wide as seismic, full waveform inversion algorithms. So you've got this real thing of very good, experienced people with incredible horsepower in compute, much better than anything in history, along with dramatic technology steps from the service providers. And then I think the magic we have right now is the team is very engaged on the digital side and very engaged with using the technology and the AI to test new theories and see what else is there. So that's a little bit about the magic. Is it repeatable? I'm never going to say that with exploration. My father was a geologist, and I know you curse yourself if you say that. So I don't think I'd necessarily bank on that. But we've certainly had a good year. We have some very good prospects next year. And as far as increasing capital in the space, the lesson for life from us is always quality through choice. Create as many opportunities you can, high grade down to the very best ones, and that gives you a higher chance of success than you otherwise would. So that, to me, is what's so important is you keep quality through choice. I think we're spending around $600 million a year right now on exploration. I would not want to push that up despite the success because it forces quality. So thanks for the question. Congrats to the explorers for a great year, and we just need to remain capitally disciplined and make sure that we're pursuing only the very best opportunities. Thanks, Paul.

Thank you, Paul. We will go to Michele at Goldman Sachs next.

Congratulations again on the strong delivery this quarter. I wanted to come back to the CapEx budget. So you reiterated the guidance for this year, and you've got a relatively wide range for '26, '27 of 13% to 15%. I was wondering, in an uncertain macro environment, if you were forced or decided to go to the low end of that range, where would you find the levers of flexibility to lower the budget effectively from the 14.5% of this year? I find it's an interesting time to start to think about some of those moving parts.

Kate, why don't you take that one?

Yes, I will. Thank you. Michele, yes, so we have got a decent range around the frame for the next couple of years. So that gives us plenty of space to maneuver, I think, in different price environments. As you look at this year, we've guided to around 14.5%. If you take out of that the final bullet on our BP Bioenergy and organic, then you're actually sub 14% on an organic basis. If prices were to dip, we've got plenty of opportunity to take ourselves down to the bottom of that frame. And we'll continue to be very careful as we deploy every dollar if prices are strong and we choose that we actually want to drift up towards the top of that frame. I don't see any need for us to let go of the tight discipline that Murray and I have put around capital. I think it forces the right conversations in terms of the value and returns focuses that we're pushing into every investment decision that we are now stepping through. And you can -- you've heard us talk about this on previous calls that, that discipline and that approach to our capital investment is so critical to us. As we seek to drive improvement in the operating cash flow going forward and making sure that we're choosing the very, very best of the opportunities at our disposal. We've got probably one of the richest sources of opportunities to consider that we've had for a very long time right now, which is a great position to be in. And I'm very comfortable with the range and the flexibility that we've got, and we've talked before where we would go if we needed to take ourselves down to the bottom of that range, and there's plenty of opportunities around some of the onshore drilling, which we could choose to slow down. There's a little bit around the exploration playing at the edges, depending on how much of our rigs are committed over the next 12 months. But we have space and we have flexibility within that 13 to 15.

We're going to take the next question from Henry Tarr at Berenberg.

I wanted to ask about Iraq. Can you give us any more details on the sort of economics for BP of the contract in Kirkuk? And then obviously, others have entered into the country and there's a sort of large program planned. How material from a macro perspective, do you think the overall impact could be for production growth in Iraq if we look out sort of 3 to 5 years?

Thanks, Henry. I have to be careful on economics. The nation has not yet published the production sharing agreement. And until they do that, it's very difficult for me to say anything under the restrictions that we have. What I will say is progress since I last talked to you. We've done the initial production test and agreed that with the nation. That's 328 kbd of black oil is being produced. And the teams on the ground now, 45 people on the ground in Kirkuk, starting to work on well work jobs, acid jobs, compressor rewheels, getting procurement contracts in place, et cetera. And we look forward to helping the nation ramp up that field over time. I think the stuff that I can say on the terms are -- they're obviously better than the first-round terms. We're on round 8, and each round has been incremental is my understanding across time based on what's been published publicly. And we do have price upside in this one. We do have the ability to take price on gas as well, which has not happened in previous rounds. We have exploration rights on the acreage as well, both surrounding and deeper. So it's a much better enhanced contract than we saw in the Phase 1 terms of Rumaila, kind of, gosh, how many years on is that now, almost 20 years on. So that's probably all I can say about the commercial terms of Kirkuk, but we're very happy with it. And in due course, when we're allowed, we'll happily share the details. with the marketplace. I think on the overall capacity for Iraq, there is a lot of oil there. And obviously, we've seen a few other deals being signed recently. And I guess my response is it's what the world needs. We continue to see oil demand moving forward strongly. We see strong demand for that oil. We perceive that some of the non-OPEC Plus is pretty much tapped out after February, March, April next year, and then we see flat to declining production outside of OPEC Plus. So it's going to be dependent on places like Iraq to help fill the demand that's coming forward. So I think the world is going to need it, but it wouldn't be right for me to talk into Iraq's production capacity. That's something that the nation will have to talk about as opposed to myself. Hope that helps, Henry.

We're going to take the next question from Kim Fustier at HSBC.

I wanted to ask about the Venture Global case. You've won the LNG arbitration case unlike one of your peers. Why do you think your case was successful? And when do you think you might receive the $1 billion of damages that you've asked for?

Yes, Kim, I'm obviously not going to comment on any other cases. I'm not familiar with them, and it would be inappropriate for me to comment on that. As far as our case goes, we're very pleased with the result. Congratulations to our lawyers and our traders for having achieved this. The next phase on damages is being organized with the arbitration panel. A date has not yet been set. I'm sure there will be an update when that occurs. And as far as the damages themselves, that's not a number that is our number. That's -- we don't recognize that number. So all I'd say is we're pleased. We look forward to the next stage. We'll update you when we're aware of when that's happening. And I'm very pleased with the result from the arbitration. Congrats to the team.

We're going to go to Josh Stone at UBS, please.

A question for Kate on the balance sheet. I'm curious as to how much attention you're paying to your gearing ratio on either a net debt to capital or equity basis because the reason I ask is as you get more of these cash proceeds in from asset sales, you will -- you're effectively selling parts of BP, your asset base will be coming lower and that's also before the impact of impairments. So maybe just talk about how you're thinking about these ratios because I appreciate you've got like an absolute net debt target, but I think the gearing ratio is also relevant here. So maybe some comments on that would be helpful.

Yes. Josh, thank you for the question. Let me step through how we think about our balance sheet because I think if you just bear with me, and I'll take you through my thinking because I think it's quite important context. So financial resilience is really important to us as an organization as we move forward. It allows us to execute on the opportunities that we have as they present themselves. And it's comprised of a number of things. And the first thing that everyone can measure us against is net debt, and we've now put a target against a material reduction in net debt by the end of 2027, and we've put a $14 billion to $18 billion, which we will deliver. If you think about where we stand today at '26, that would be a $10 billion reduction in terms of the net debt stack. But of course, I think if you remember some of the slides that I used to talk about balance sheet and financial resilience at the Capital Markets Day in February, I was trying to be pretty transparent that we understand our total liabilities and the drain on our operating cash flow, those type of commitments is not just around debt. If I think about some of the other big components, we have over $1 billion a year going out on Deepwater Horizon. So another 2 of those will go before the end of 2027. So that's $2.2 billion. We've got a level of prefinancing of the '26 hybrid I referred to earlier, that's $1.4 billion. So even if we do nothing else, where we stand right now, our liability stack will reduce over the next 2 and a bit years by $13 billion to $14 billion. And that's how we think about it as opposed to contemplating gearing. We haven't got a gearing target. We've got a target on net debt. That's the first priority. That's what we will deliver. But I hope you can hear from my language that we think about the totality of our liabilities, and we're cognizant on the total cost of all of those.

We'll take the next question from Jeff in TPH, please, back over in the U.S.

We were hoping to also ask about the structural cost improvements, which look to be progressing quite well, especially based on the supplement disclosure, the roughly $400 million improvement quarter-on-quarter there. But I'll actually gear my loan question here to follow up on BPX, if you could dig into basin-specific plans a bit more and maybe give us a sense for how you plan to pace activity adds in the Haynesville specifically over the next 12 to 18 months or so and maybe how the Eagle Ford may play a role, if at all, as part of that.

Yes. Great. Thanks, Jeff. Thanks for the kind words on cost progress. I'm sure Kate would like to update somebody if they want to ask a question on that one. As far as BPX, our plans in the Permian as we built out the infrastructure, we, of course, want to keep that full now that we've built that out. So 2 to 3 rigs to continue to keep that full for time. In the Eagle Ford, we continue to be very excited with the downspacing inside the oil window of the Blackhawk and the refrac programs. The downspacing wells are doing better than the motherbore than the original wells simply because fracking technology has moved on so much from when they were drilled a decade ago. And the refracs, similarly, we're seeing much higher production on refracs than we did in the original wells from a decade ago that Petrohawk would have drilled. So those are places that we'll continue to push and push the liquids side over time. And then on the gassier window, of course, we've got the associated gas from the Permian. But equally, we have a fantastic Hawkville gas, which is in the Eagle Ford, and we have fantastic Haynesville positions as well, the core of the core. On the Haynesville itself, we'll follow the infrastructure is the way to think about it. As I said earlier, the teams have been doing a fantastic job on driving capital efficiency inside that basin, setting record after record on production capacity from the wells now up to 80 million a day on this latest 4-mile horizontal. And what we -- through our trading and marketing organization, we've been busy establishing offtake points. So we'll just gradually continue to grow the Haynesville in line with the infrastructure build-out, really infield gathering rather than any main export issues. And we'll be contemplating 2 versus 3 rigs as we head into 2020 -- into the fourth quarter -- into the end of the fourth quarter and into 2026, and we'll update you from there. But tremendous resource, tremendous performance by the team and good gas prices, obviously, as well that we hedge out, and we look forward to growing that part of the business. I hope that helps.

We're going to go to Alice at Morgan Stanley. Alice, I know you're deputizing for Martijn, who's also here in Abu Dhabi. Over to you, Alice.

I have a question about downstream. So you printed a pretty strong results sequentially, but also with a number of moving parts. So of course, there was the successful delivery of the cost reductions, but also supportive macro for refining and then on the other hand, weak trading. So could you please give some insight into the contribution of each of those elements? And then on balance, what could we expect the run rate to look like?

Kate, over to you.

Yes. Thank you. Alice, let me try and break out the components of the improvement in the downstream. I would say it's been a 9-month period of really good performance across pretty much all of the business, actually in terms of the way that we've seen the organic improvement coming through, firstly, on the customer side, a number of things. We've seen improvements, I would say, in almost every area of the customer side, whether it's aviation, Castrol is up 21%, I think now year-on-year for the 9 months. We've got stronger performance coming through the tight integration that we've got between fuels and midstream. That's something we've been working really hard on that's coming through. And we've got really good cost reductions. So structural cost reductions delivered for the 9 months so far inside customers is about $0.5 billion. And then the other component of the improvement in the downstream operating cash flow from customers is around the BP Bioenergy. So as you recall, we consolidated that now. So you're seeing an improvement in terms of the consolidated earnings versus just [indiscernible] about $300 million. And then if I look at the product side of it, the refining portfolio is delivering superbly now. We've got refining availability year-to-date at 96.4%. That compares to the 96% that we set ourselves as a target back in February. That's a result of conscious investment and systematic improvement in the maintenance and integrity of our kit. And as a consequence, it's running well. And as the refining margin improves, as it has done in the last quarter, we're able to capture the maximum of that. I would also say that refining have done pretty well on their business improvement program as well in terms of reducing their costs. They've reduced their cost by about $200 million further 9 months. And then finally, perhaps on trading. Trading had a weaker quarter this quarter, but they had a very strong quarter in 2Q compared to others. We were very pleased with that. But as I look at the 9 months year-to-date, trading is pretty much in line with where it was last year. So very comfortable with where trading is. So that's quite a long answer. Hopefully, that's broken it down to enough detail for you to be able to follow the various component parts, Alice.

We're going to take the next question from Peter Low at Rothschild Redburn.

Maybe one just on the Gulf of America. Now that you've taken FID on the Tiber-Guadalupe project, does that open the door to a potential farm down of your Paleogene positions? Or what's your current thinking on the optimum time to do that kind of within the development of those assets?

Yes. Thanks, Peter. Yes, very, very happy to have taken sanction on Tiber. It's obviously the second sanction inside the Paleogene, Kaskida a year ago and now Tiber, 280 kbd boats. We own them 100% with tremendous resource recovery potential sanctioned and potential moving forward. And I was really pleased with the projects team. They were able to knock $3 a barrel off the development cost on Tiber by effectively photocopying what we've done on Kaskida. So build -- design one, build many. So that's all very good. We are in conversations with counterparts about the potential farm down in the Paleogene, and we'll do this for value. That's all that we have in our minds is how do we do this for value. And we want to make sure that it's accretive and that it's in the shareholders' interest to do that. But we continue the conversations -- and like all other divestments, we'll update you when we have something to tell you. Thanks for the question, Peter.

Thanks, Peter. We'll turn to Mark Wilson at Jefferies.

I will bring it back to Bumerangue. Again, still got a lot of data you mentioned and that appraisal will take a flow test. The release a few days ago spoke to an early production system. It sounds to me like a flow test there would have to be for a prolonged period of time to test multiple areas of a large column and fully understand the CO2 mix. That also sounds quite costly within a $600 million exploration budget if that includes appraisal. So first, I'd like to ask if I'm visualizing that work scope correct for, say, 2027 in terms of what flow testing is needed? And would you appraise that at 100%? Or would we expect overall exploration cost to go higher to accommodate the Bumerangue appraisal?

Yes. Great question. It's a pretty good reservoir. So we think the flow test, where we're drilling the second appraisal well will give us a pretty strong indication of what the rest of the reservoir will perform like. We, of course, could be surprised as we go through that. But given the strength of the seismic, what we're seeing on the logs, we think that is the case. As far as -- so we'll do that somewhere around 4Q '26, early 2027. And we do have a team working in early production scheme. Of course, it will depend on how the flow test goes. The flow test is really focused on productivity of the wells, to be honest, and how many wells we're going to need to drill. That's the primary focus that we'll have on that as we've done all the sampling in the sidewall core already from the initial appraisal well. So it's mainly about how many wells we need to produce the reservoir over time. As far as timing of partnership, that's something in time, we will bring in a partner for sure. You probably don't want to do it until you're through the appraisal well and the flow test because that will have an awful lot more information that's derisked. But that's, of course, a decision that we'll think about with the Board as we move forward. And the exploration, I'm not quoting an exploration number, including appraisal at this stage. We're somewhere around $500 million or $600 million on exploration. We're drilling about 15 wells a year right now, and we'll update you as we work our way through this as we enter '26 and '27. But we will stay inside that $13 billion to $15 billion capital frame that Kate talked about. That's very important. So I hope that helps, Mark.

We're going to go to Bertrand at Kepler next, please.

Yes. Coming back on the Venture Global arbitration. Murray, you've just mentioned that the $1 billion plus in damages that were in the press was not your numbers. Can you elaborate a bit? Or are you seeking a higher number?

Bertrand, thank you for the question. Look, this is -- you're quoting a number that was in a press release from Venture Global. We have not disclosed anything to the public markets around our viewpoint on this. And as it's a commercial process, I cannot disclose anything because it could impact the arbitration process, and I'm not going to do that. So I'm afraid I'm just going to have to say that, that was their number, not ours. And in due course, we'll file our claims with the arbitration panel. And when the arbitration panel decides, they could make their views public. But I have to be very careful in the process, and I can't talk about anything commercially. Sorry, Bertrand.

We're going to move to the U.S. again, Jason Gabelman at TD Cowen.

I wanted to ask just on the equity affiliate portion, given you have quite a few of them at this point. And as you think about what the overall net contribution of those affiliates to your cash flow is, I'm wondering if that's changed at all given the JERA Nex BP joint venture and Beacon Wind within that joint venture being canceled and perhaps less cash infusions into that joint venture, but conversely, the success at Azule with the Namibia explorations resulting in perhaps less cash distributions from that entity in the near term? And just how that kind of rolls up into your overall views on distributions moving forward.

Yes. So shall I -- I'll take that, if you like. In terms of JERA Nex, the way to think about that is it's about creating for us in the future optionality, but in a very, very capital-light way. So JERA Nex will make their own decisions in terms of the projects that they execute and the sort of hurdles that they're testing against. But from our perspective, it will be very capital-light and capital that we -- if we were required to put capital, it's going to have to compete with the other calls on capital in our portfolio, which is a pretty high hurdle. In terms of Azule, a very different type of joint venture. It's been a very good quality joint venture for us so far. I think we've got about $7 billion of distributions from it year-to-date -- sorry, in total since inception. It's now self-funded. It's got a PXF and it's also issued its first bonds externally. So in terms of its being able to finance itself going forward and its growth, that's how we think about it right now. It has the ability to do more with regard to external financing. So we're not expecting it to be a drain on our capital. Of course, to the extent that it is recycling its own cash flow to invest in opportunities like Namibia, then you would see a slight reduction in terms of the dividends that we receive from that organization, but it's too early to be able to scale that for you.

Yes. And if I just added a few things on Azule, we don't often talk about it, but it's had tremendous success, Jason. Agogo came online earlier this year, 8 months ahead of schedule. So congratulations to the team for doing that. MGC is the next major project that's going to come online shortly. And they, of course, have the exploration discovery near the LNG plant of TCF and a couple of hundred million barrels of associated condensate. So in Angola, it's doing fantastic. I was down there recently to celebrate the Agogo start-up and Gordon was offshore on it. So just a tremendous joint venture with Eni that's doing very, very well for us, and we're very pleased to have expanded that into Namibia and the success we're seeing in Namibia. So I hope that helps, Jason.

I think we are probably at the final question given time. We're going to take that from Maurizio Carulli at Quilter Cheviot.

Well done for the positive results. Can I have a bit more color on the 20% increase in Castrol earnings and what has driven it? And also, if I may squeeze in an additional question. Is it possible to have more detail on your recent strategic investment in the electronic cooling solutions?

So maybe you want me to take that one. Yes. So this is a consequence of very deliberate progress that Michelle has been executing now for -- I think this is the ninth quarter where we've seen quarter-on-quarter progress. They have a strategy of onward upward forward, and it's deliberately seeking to make their organization as cost-competitive as it can be and grow volumes, which they've managed to do systematically over the last couple of years in almost every part of the business. The other part of the improved delivery in Castrol, of course, is a consequence of the fluctuations that we've seen in base oil and additives and they hit a high post-COVID, those have tapered off a little bit. So that's also coming through, which is helping. But it's about very deliberately growing volumes, driving costs down to improve their overall operating cash flow delivery inside the organization. So that's doing really well. Do you want to talk about the liquid?

Yes. The liquid cooling for data centers is an interesting opportunity. They've signed a couple of deals with counterparts. It's commercially sensitive, so I can't name the names, and they're in trial on that with a few companies. It's a long-term growth potential for the business. that looks quite interesting. It's quite a competitive space, and we -- we're hopeful that, that starts to develop and at a faster pace moving forward. But thanks very much for the question, Maurizio. Nice to hear your voice.

We are going to finish promptly on the hour. Irene, Chris, I know you are still pulling. Maybe please follow up with the IR team in London. Happy to do so. A big thanks on behalf of Murray, Kate, myself, thank you for listening. Thank you for the continued interest in BP's results today, and we'll stop the call there. Thank you again.