Hindalco Industries Limited analyst Q&A

Thank you very much. We will now begin with the question and answer session. Our first question comes from the line of Sumangal Nevatia from Kotak Securities.

So my first question is a few clarifications. So sir, when we say hedging 31% for '26, now are we implying for the second half, say, 31% of the volumes we will sell in the second half or it's I mean, how can you just explain that?

Yes, let me explain. So Q3, it's 31% at 2,700. Q4, because of the recent LME run-up, we have been adding to our positions. We are now at 49% at 2,760. So on top of our previous positions, as the LME has gone up to 29%, we have been adding to our position. So Q3, we have had 31% at 2,700. Q4, we have now 49% at 2,760.

Okay. And sir, how is FY '27? Have you started locking in '27 volumes also?

Yes, we have started to lock in FY '27. We currently have roughly 10% at 2,800.

Understood. Sir, my second question is on Novelis. Now one clarification, which we're not able to understand. What is our net impact on the tariffs on the earnings of Novelis? Assuming today's tariffs, today's scrap spreads, here, I'm not asking for any speculative future scenarios of tariff reversals or something, but just want to understand on today's scenario, adjusting for all the lead and lags, which we kind of witness, what is the net impact?

Yes. Dev, you want to take that?

Yes. So basically, we said that this quarter impact is $54 million. Remember, we had guided to $60 million net impact of tariffs at our last quarter earnings call. So the second quarter came at $54 million. The first quarter was at $28 million. It was a part of the quarter impact. Now we have already been telling you that we have a mitigation strategy in place with relocating more getting access and relocating more manufacturing in the U.S. versus Canada and that will help us to continue mitigating more impact to the extent that towards the end of the year, towards the end of this fiscal, a very significant part of the mitigation would have happened, Sumangal. So that is really what it is.

And Dev, just to understand, are we factoring in the benefit of the scrap spreads in the North America because of these tariffs in this net impact or that scrap spread is outside of the net impact?

The scrap spreads per se is outside of this impact. We don't factor that in. We do factor the benefit of the higher Midwest Premium, both in terms of getting the arbitrage. And also, there is a factor of scrap benefit just because of the elevated Midwest. So that's the way we do it.

Okay. I understand. My next question is just one last one on, overall, our leverage and what is the peak level of debt we are estimating over the next 2, 3 years at Hindalco console level. Just want to understand, given that we are seeing various cost overruns at Novelis and till today, we have just spent around 50% of the Bay Minette project capex. And then at the same time, we are taking up more expansions in India. So as per our budgeting estimates, I mean, what is the net level of debt we are looking at in terms of peak?

So Sumangal, if you remember, in April, when we did the investor call, we committed that the console net debt to EBITDA will not cross two. So what we are doing now is as we do any of the India capex announcements, we are taking the Bay Minette number of 5 is already in our calculation. So all I can tell you is that we will not cross 2 in the next 4 years. We will make sure that the console net debt to EBITDA is kept below 2. And our Indian capex is what we have announced already, that is already in our calculation. And we will make sure that we will phase the capex so that we will as on a console level, we won't cross 2.

Understood. And just one accounting clarification. There's a lot of intersegment profit and unallocated income expense, which where there's a lot of swing every quarter. So I want to understand, are these noncash items or cash items?

No, Sumangal, I guess you're talking about the India business, right?

Yes.

The only swing is when we sell metal from upstream to the downstream. So that is the only thing that happens because just like Novelis has, when you sell upstream to downstream, the metal sometimes if it does not get sold, then it will come up in that unallocable because it's in the inventory of the downstream. Whereas on the upstream side, we have to take it as a metal sold and book the profit. So that's the only correction that we do in that annnexure side for you to get that number right. Otherwise, there's not any other things going on there.

Our next question comes from the line of Satyadeep Jain from Ambit Capital.

Mr. Pai, just wanted to understand the thought process on the doubling down of aluminium smelter. So on one side, you are hedging the second half exposure, you're increasing hedging, which means which sends a signal that you're cautious on this study. And on the other hand, you're doubling down on the smelter, which seems to suggest you're positive. A very conflicting message. I'm just trying to understand the thought process. And tied to that would be, when you're looking at expansion in India smelter, did you consider recycling at these spreads, the thought between recycling versus smelter in terms of expansion, just on the this expansion, just trying to understand the thought process?

Yes. There's a lot of points you're covering. So the first thing is the hedging part, if you remember, our strategy has not changed. Every year, we try to do about 20% as an insurance hedge for the next year. So say, roughly by December, January, we would like to be at 20% for FY '27. So that's an insurance hedge, and it has not changed. As we get closer quarter-on-quarter, if there is a run-up like we see now, then we add because we want to lock in Q3 and Q4 numbers. Now the smelter expansion is a much longer-term strategy based on the EBITDA and the LME projection. So we are fairly confident now that our cost curve will remain with our coal mines coming in at the first quartile. We are fairly confident that the LME will at the lower end of the LME, and I've said this many times, is now going to be more like 2,200 to 2,400. And at that level, when we budget, we can see a very good return. Now if LME continues to be strong, then it is very good. But it's the fact that we have coal and bauxite backward integrated makes us go into the smelter expansion. And the third part is we are expanding downstream in India. So as we expand that downstream, you see we already reached 600 Kt of downstream capacity. And we are aiming towards an EBITDA per ton of $300 per ton and higher. So all these put together gives our strategy of leveraging the upstream because of the integration strength and then expanding the downstream as well to take into account as the Indian economy grows, more manufacturing comes to India. So a long answer, but it's a complex question you asked.

And on between recycling and smelter, the entire capacity expansion?

Yes. So recycling, we are also going into recycling. So the 50 Kt recycling on Aditya boundary has already started. So we are committed to get the recycling volumes up to 100 Kt in the next few years. The copper recycling plant will get commissioned next calendar year. So even in copper, we'll have 50 kt of recycling. So we think that recycling and primary both will have a role to play going forward. Recycling is where India has to catch up a lot, and we are actually going to be leaders in both copper and aluminium recycling.

Secondly, on the de-risking of the project in the context of Novelis also, initially when the cost overrun happened, I mean, we assume that this is engineering, there was some, I think, flow over or I'm not sure which one, something on the engineering design. But again, that has come up as 50% of the cost inflation this time is again because of engineering. So what has really happened that this happened the second time the same reason? And how do you look at de-risking the India projects? I know that's maybe you're going into Alabama was somewhat greenfield. This would be back in your own backyard. But just what really surprise happened the second time, which you already had figured out the first time? And then how do you de-risk the India projects, large capex happening there?

Yes. I'll let Steve take the first part, and I'll take the second part. Steve, go ahead.

Yes. So thank you, Satish, and thanks for the question. With the advancement of Bay Minette, we've made great progress over the past 12 months at the site. We're about to commission the cold mill next quarter and then the rest of the equipment through the remainder of next year. The strategic nature of this project in ensuring that we're ready to continue to expand, as Satish talked about, because of the situation in the U.S. being very significantly short of capacity for beverage packaging, for auto and specialty products. And then with the tariff situation, this project becomes even more strategic. And so the elements of the increase from roughly 4 to around 5 is 50% of it is inflation, standard inflation as it relates to some of the equipment and materials, but mainly around contract labor because of the significant industrial construction projects going on in the U.S. So that's the big driver. And then we have recognized that there is some complexity as well associated with some parts of the facility that we were still finalizing up until just this quarter. And so as we saw that, that also increased the cost, and that was roughly probably 30% to 40% of the additional increase. And then the remainder is ensuring that the site and plant is ready to be quickly expanded in the event that we are successful in some U.S. tariff relief with the administration and so that we can continue to catch the significant growth that we see over the next 5 years in the U.S. market.

So just wanted to check on this only. I know when the entire inflation happened in the first instance to $4.1 billion, we would have thought that the entire on the engineering side would have been de-risked. So maybe taking the cost inflation bit out. But again, what is the complexity you saw in the final stages that the $4.1 billion would have already accounted for a lot because that was a time when you saw the engineering issues. What is something that even you could not see and these issues cropped up at the later stage? Just trying to understand especially that engineering issue complexity that happened the second time?

Yes. So as we talked 1.5 years ago about the increase to the $4 billion number, we said we had a significant portion of engineering completed. We did not have 100% of engineering completed. And so as we continue to work towards getting that fully completed. That's where we saw some differences in quantity and variances occurring, whether it be in amount of cement, amount of steel and other materials or amount of hours required from contract labor to construct the plant. Now we are 100% complete in engineering. And so we are confident in the $5 billion number as we sit here today.

Okay. And maybe Satish on India, if you can.

Yes. Let me take the India question now. So I guess we have unfortunately learned many lessons along the road. So we have set up a very strong project team and a project monitoring team for all the India projects that I'm personally sort of reviewing regularly. But taking Steve's point, we have hired quite a few people on the civil electrical side who are looking at many of these engineering drawings and the conversion of these engineering drawings to release for construction type where the cost inflation or the numbers go up. So we are trying to get to a high degree of maturity on the conversion of engineering drawings to the release for construction for the India projects to not fall into the same issue. So a lot of hiring of experienced people from the industry with civil, electrical because it does not normally come from our equipment. A lot of these things come from the civil side. So we are plugging the gap to make sure that the Indian projects do not have this issue.

Our next question comes from the line of Amit Lahoti from Emkay.

My first question is on Novelis. So with the implementation of tariffs in the U.S., clearly, the capex intensity and the cost of doing business has gone up. And we have 420,000 tons of volumes contracted for $1,000 margin. So for the remaining 180,000 tons, can we seek higher margin to pass through the tariff and capex inflation?

The answer is yes, but I'll let Steve say the yes. Go ahead, Steve.

Yes. No, it's a great question. And obviously, with the tight market in North America and the tariff regime that's been put on, especially with the 232 aluminium tariffs at 50%, there will be opportunities to see price increases on uncontracted and ultimately likely when those contracts come up for renewal as well. So I think the overall price outlook in North America is to the upside.

Okay. And then my second question, again on hedging. So see, in a way, the market waits for a good cycle for a long time. And when it actually comes, the companies are found on the short side of it through hedging. So how should one think about the investment case given that the upside is deliberately capped, how do we think about it?

So no, I don't think that you should look at this deliberately capped because the LME is very volatile and can go up and down. So if you take a look at it, I think that when you hit $2,800 levels in the short term, it becomes, I think, for us, very important to lock in that because you see Q4 is a quarter where normally the Chinese shutdown happens because of New Year and LME can come off a few hundred dollars. So which is why we think we took that up in the Q4 because Q4, we want to lock in the numbers to have a strong quarter. So in the long term, as I repeat again, we do only 20% to 25% and keep 75% open. In the short term, when we get very close, we try to catch the thing going up.

Our next question comes from the line of Indrajit Agarwal with CLSA.

I have two questions. First, Mr. Pai, if you can explain what is happening in the MJP premium market, while LME is holding up much better, MJP has kind of collapsed. So what exactly is driving this weakness?

So you see the difference in the aluminium versus copper is that the regional issues get resolved in the regional premium. So on one hand, MJP is low. Because of the tariffs, the Japanese, the Korean auto companies got impacted. in Q2. So the aluminium demand in Asia was low, especially East Asia. So the MJP is sub 100. But then you switch to the U.S. because of the tariffs, the Midwest is running at $1,600, $1,700. But the LME remains the price across the world the same. So what you're seeing in MJP is a reflection of the aluminium demand. Now that South Korea and Japan have signed deals with the U.S., we are assuming that in Q3 and Q4, the MJP will start to go above 100 where it should normally be.

Sure. Second question is to Steve. If you can remind us again the beverage can sheet market in U.S. So what is the demand? What is the import like? And would there be enough demand for the additional 600 kt Phase 2 that we are planning at Bay Minette?

Yes. Sure. So the overall beverage packaging market globally is very strong. We see that growth at 3% to 4% over the next 5 years on a CAGR basis. In the U.S., it's extremely strong as well. And we're starting from a base of being short can sheet in the U.S. market by somewhere between 400 to 500 kt. So alone, as with the tariff situation, that will all get domesticated between Bay Minette and the new ADI mill as well. From there, we see strong growth and ability that as we get into 2030 that the market will be short again, not just beverage packaging, but clearly in automotive as well and some specialty products. And this mill would have the capability to serve all 3 of those. I hope that helps.

Our next question comes from the line of Pinakin from HSBC.

My first question is that you mentioned in your opening comments that to do the capex, you would want to keep the net debt to EBITDA at 2x or below 2x. And the India capex will be the one which could get phased out. So just trying to understand that if over the next 1 to 2 years, profitability at Novelis does not improve for whatever reason, should we expect a delay in the India project commissioning time lines?

So first, welcome back, Pinakin. So just to answer your question, roughly the projects that have already been announced will continue. The ones that we will look to phase are things that we have not yet taken a call on such as the Mahan 360 Kt smelter expansion. So the ones that we have already announced are already booked into the projection to say, and we are comfortable we'll stay below two.

Got it. My second question is, sir, if you go to '27 and where you have given the India project expansion schedule, can you give us a sense of the India business capex next year and thereafter, including the $750 million of equity infusion that is to be done at Novelis?

Sorry, the question was in FY '27, what is the capex?

'26, '27 and '28. How should we look at the capex given the projects that are on committed to in India and the $750 million of equity infusion at Novelis?

So Pinakin, let's take the capex part first. So this year, we have said the capex is going to be around INR8,500 crores. I think that next year, we are expecting around INR11,000 crores. I think that is FY '27. FY '28, stay with me because I'll have to give you a number maybe next year. But this year, about INR8,500 crores, next year about INR11,000 crores is what we have put into our projection. The INR750 crores that we would raise at the AV Mineral level for the equity infusion has also been put into our calculation when we talk about the consolidated net debt to EBITDA staying below two. Because anyway, the net debt to EBITDA is a console number.

Our next question comes from the line of Ritesh Shah from Investec.

A couple of questions. Sir, you indicated that our intent is to cover up for the cost of capital at Novelis. Sir, what is the underlying assumption on the cost of capital that we are taking in? That's one. Second is you did indicate about the second phase of expansion. Any time lines to that, which will help us improve the return ratio profile? And third, related question, what's the rationale of infusing $750 million from India if one had to take into account the cost of debt and cost of equity?

Yes. Maybe before we go to the Bay Minette cost of capital and IRR to Dev, the $750 million, we are going to raise as debt at the AV Mineral level and put it as equity into Novelis because a couple of things. I think that the strength of the Hindalco balance sheet means that we can raise money on the Hindalco side at a much more attractive cost. Number two, we don't want the net debt of EBITDA of Novelis to go too high. We have bond owners that we are worried about. We don't want the ratings of Novelis to get negatively impacted. And fourthly, we believe that showing parental support at this time would be a strong positive for Novelis. So which is why for a variety of reasons, we want to raise the debt at the Hindalco AV Minerals and put it as equity into Novelis. I'll let Dev now answer the IRR, cost of capital for Novelis and the Bay Minette project. Dev?

Yes. So Ritesh, so typically, as a policy, we don't give out the WACC, but let me give you enough inputs to at least make you comfortable on the intent of your question. So you know all the debt raises that we have done. I mean, our cost of debt in the U.S. is somewhere in the 5 handle, low 5 handle consolidated cost of debt. The point I'm making is that the cost of capital in the U.S. is very efficient. Now I understand, I mean, in the WACC, you'll have the cost of equity, but still cost of capital is very reasonable. And then what we have told you is that even with the Bay Minette overrun, we will not be exactly on double-digit IRR, but a wee bit below double-digit IRR. So again, I'm just asserting the point that we have a safe margin versus the cost of capital when it comes to Bay Minette. And I mean, as we have assessed the project cost of Bay Minette, we have also assessed opportunities and learning from all the things that we have done on cost efficiency projects, we discovered that there was even more opportunity in Bay Minette for cost optimization. There are pricing opportunities, which was already alluded to earlier by Steve. So all in all, we are discovering more efficiency opportunities even in Phase 1 of Bay Minette. So it makes us feel comfortable that besides the strategic rationale, even the financial rationale of Bay Minette is fairly intact even with the first phase.

Sorry to harp on this. But again, like I think when we indicate cost of debt closer to 5% or less than 5%, then probably we don't...

I said it's a 5 handle. Yes, go ahead.

Yes. Then we don't need the $1,000 EBITDA per ton number because it more than adequately covers up for the cost of capital because if I just put 60-40 debt equity, if I assume cost of equity at 11%, the number comes closer to 6.5%, assuming, say, 5% handle. Then how should we put the 2 variables into the equation? One is when we say covering up for the cost of capital and the number of $1,000?

Yes. I mean well, I don't know how where you're going with this, but you are just on the point that the cost of capital is indeed very efficient, probably a little bit above the WACC that you have calculated. But the point that we are saying is exactly what you're also getting to that the $1,000 and now with cost efficiencies that I spoke about, it is not just $1,000. It is decently above $1,000 with pricing and cost. So essentially, that's the whole point I'm making that there is enough spread between the WACC and the IRR that we are getting based upon all the updated assessments that we have done, Ritesh.

Sure. That's helpful. Just a small clarification to what Mr. Pai indicated, raising debt at AV Minerals, I think the last bond issuance, what we had at Novelis was at around 6.375% coupon. So when you say something at Hindalco AV Minerals, at attractive debt, what is the coupon what we are hinting at because...

It will be lower than that. My treasury guy is waving his hand, don't give a number, but it will be lower than that.

That's great. Just a second question on Europe, probably Steve can take it up. Sir, how are we looking at the impact of CBAM in Europe? Specifically, if you could highlight our sourcing of scrap within Europe? And if you're sourcing any metal from outside Europe, what would it mean to the premiums and the underlying economics for the European operations?

Dev, do you want to take that? This is scrap.

Yes. So I can surely talk about the scrap and we have enough scrap availability in Europe to be able to take care of all our needs. We don't depend upon imported scrap in Europe. In fact, as a policy, we are kind of encouraging the regulators to think about not letting scrap go out of Europe because there is some scrap leakage that happens. So Europe is self-sufficient. And so we have no concerns on that. We would like to sort of see if the policymakers would react positively to not letting scrap go out and let the scrap remain in Europe. And then that indirectly also addresses your CBAM question, I guess.

Our next question comes from the line of Somaiah V from Avendus Spark.

Sir, first question is on the India business. On a Q-o-Q basis, the EBITDA per ton increase of close to $40, $50. This seems to be lower when compared to the LME increase. If you could just help us understand the bridge here, impact of cost of production of the review premium that you are referring to earlier?

So look, the EBITDA per ton, you also have to take into account the alumina sales that we do where the index has come down. So that has got an impact. The second thing is Q2 coal costs, while they have been attractive, it's a monsoon quarter. So coal costs Q-on-Q, meaning sequentially are higher. So the cost of production in the Q2 level, I was expecting that question really is about 3% to 4% higher than Q1. When you do the math, we have had a positive write-back of the RPO obligations that we have put in. But if I take that out, the cost of production has been about 4% higher in this monsoon quarter. So alumina and the cost of coal going up is why the cost the margins would not have been as high as just the straight LME going up. And the other part, of course, is also the regional, meaning the India premium that we realized also has been slightly lower in Q2 versus Q1.

So on that note, what is the expectation in terms of cost of production for the current quarter, Q3?

The Q3 should be flattish to 1% is what we are assuming. Coal costs will come back, but CP coke prices have gone up a bit. We do have some planned shutdowns. So it should be flat to 1% in Q3 versus Q2.

On Novelis, I just want to understand the scrap-led benefits in Q2. Obviously, from the industry pricing that we track, it seems to be a very sharp improvement in scrap spreads between Q1 and Q2. Scrap prices not going up to the levels of NWP increase. So is the full impact of the scrap spreads already shown up in Q2 or if there is a lag impact. So what would be the drivers? If you could just help us understand the nature of the contracts. Are contracts linked purely to LME as a percentage of LME or there's an LME plus Midwest Premium angle? Some color on the scrap spread impact or what one should expect in Q3 versus Q4?

Dev, do you want to take that?

Yes, I will take that question. All right. So one, I think you have captured it very well. I mean, we have not yet seen the full upside of the rise in Midwest because it kept rising during our September quarter. And so there is some more benefit to come. Also because of the fact that we have some contracted volumes from contracts that were locked in earlier, we were not able to take the full benefit of the falling spreads. So there is, in short, more benefit to come. And this is a wider point that I want to make to everybody who is listening that the underlying positivity in the business is becoming better. Unfortunately, we are having extraneous events, whether it is the customer facing a shutdown in Europe or Oswego fire and so on. But when you take this out and you think about the underlying health of the business, things are really looking pretty positive. Now so there's more to come. That's the short answer. Now just keep in mind also that while Midwest is doing very well and the scrap availability and spreads are benefiting, in the non-U.S. geographies, there has been some softening of premiums, particularly in Asia, where MJP is very soft. So some of the benefits of the Midwest do get taken away. But net-net, the positivity of Midwest takes over all the others, and it's a net positive, and there is more to come.

Sir, one clarification there. So I just want to get this part a bit more clearer. So in terms of sourcing of contracts, I understand we would have had some long-term contracts in place where we would have had a higher scrap price compared to what probably spot is. If that is the case, you might let us know that is one, which could lead to a lag impact. But from getting the benefits of Midwest, it's more with passing on to the end consumer. So that could be more or less captured on a spot basis. Just want to understand how the lag impact on capturing the Midwest benefits comes in the equation?

Yes. So I think we are getting into two areas. One is the metal price lag, which is below EBITDA. So that is basically in a rising premium market. you benefit because between the time you purchase and between the time you sell, the premium has gone up. So that is what is metal price lag and you would have already seen how strong the benefit from metal price lag is. Now at some point, things will go in the other direction and so on and so forth. But the point I'm making is that when Midwest or any premium goes up, as we buy as a percentage of the total aluminium cost, the dollar benefit expands, right? And that dollar benefit expansion from Midwest is what I was alluding to earlier that in a rising market, there is a bit of a lag in the expansion of the spreads, yes. And that lag will benefit us.

So sourcing contracts generally are tied as a percentage of LME or is it as a percentage of LME plus Midwest for scrap?

Yes, it is as a percentage of total BMC in the U.S., total LME plus Midwest. And remember, I mean, this is not like as cut and dry as you're making it out to be. Even the availability, there are also demand-supply factors and the availability situation in the U.S. has been getting better. And so that also creates some softening outside of just the price impact. So it's more complicated. But yes, net-net, I think you have captured what I'm trying to tell you.

Our next question comes from the line of Amit Kumar Murarka from Axis Capital.

So my first question is on alumina volumes that you booked in the quarter. Could you provide that number? And also, what's the outlook for the same in 2H?

So I think the sale for the quarter was 199 Kt in Q2. And in Q3, it will be about 170 Kt because we have a little bit of a shutdown planned in Utkal.

Okay. And then otherwise, the quarterly run rate can be taken as like broadly 200 Kt then otherwise, except the shutdown?

So we normally sell roughly about 700 to 800 Kt in a year. It's not exactly 200x4, but yes, 700 to 800.

Understood. And just quickly on the aluminium Phase 2 expansion project that you just announced. So beyond this, like which other sites will have scope for expansion for you?

See, we have 2 new smelters, Aditya and Mahan, both we had planned for 720 Kt. So both are running at 360. So Aditya, we did the first 180 is in progress. We announced the next 180. The only remaining one for expansion is Mahan, which can go up by another 360 Kt.

And that, I believe, will be taken up once the Aditya project is getting through or towards the end of that?

Sure. And just to be very clear, we are no longer like pegging or attaching the growth in aluminium capacity to renewable power as what we were doing in the past?

No. I think that we will continue our journey to try to increase as much renewable power that is practical and makes sense, but we are not going to tie it. That's fair to say.

Got it. And lastly, any update on the Chakla mine? I think it is the box cut, I think, was guided by the end of this calendar year earlier.

Yes. I think that we should be doing the box cut somewhere around December, January. That is still on track. Both Bandha and Chakla box cuts are in this coming few months.

Our next question comes from the line of Tarang Agrawal from Old Bridge Asset Management.

Three questions. One, how fungible are cash flows between Hindalco and Novelis, I mean, from a tax leakage standpoint and various FEMA regulation standpoint?

I think that this is to be taken fungibility, tax things, we look at it in the way we do. Like the example we just gave you is we can raise debt in AV Minerals and put it across. We have been doing a return of capital from Novelis to Hindalco many times without any tax leakage. We have been doing so I think that overall, this is something that we will continue to look at on a case- by-case basis.

Basically, I was coming from the point of view that given that these are two standalone businesses, but when you try to look at a console net debt to EBITDA of less than 2, it is essentially inferring that the fungibility is extremely strong. So I just wanted to verify if that's really the case?

Well, it is. I mean we are demonstrating that in these times. You're right. we look at it as one consolidated. That's how the shareholder will look at it. And then depending on strength of the balance sheet and the market conditions, we will use for financing reasons.

Got it. The second question, sir, is on the specialty Alumina business. And to the earlier participant's question where you essentially alluded that external alumina sales are in the ballpark of 700 to 800 kt per annum. Specialty capacity has been expanding. FY '25, you ended at about 560 Kt. So would it be fair to presume that 80% of those or 70% of those 800 Kt are actually in the form of specialty alumina?

No, I think you're mixing 2 things. We have a specialty Alumina business, which is based out of Belgaum and Muri where we sell about 400, 500 of specialty. What I was talking about was the excess metal grade alumina from Utkal that we sell. So these are 2 different things. The specialty alumina business out of Muri and Belgaum is, as you said, in the order of 500 Kt. We are trying to use up all of Muri to do specialty, which will take some time. Now the one part on this whole specialty alumina business, it has been doing well, but still many of the specialty aluminas are tied to the alumina API. And as you know, from the $600, $700, it has dropped to $350. So the pricing of the alumina, specialty alumina, some of our products is also impacted by the alumina price coming down, which has had an impact in Q2.

Sorry to interrupt, Tarang, sir, we had lost the audio for your line. If you can just please repeat the question once again.

Yes. So if alumina trades at about INR350 per ton, typically or 35 cents a kg, what would be the realizations in your broad basket of specialty alumina?

It's not you can't go broad basket. There are a lot of value-added products where the EBITDA per ton will be another INR20,000 above the alumina prices. So it's better to look at the specialty alumina separately. We will start to separate that out sometime next year, just like we have done downstream alumina. The Utkal alumina is metal-grade alumina sold to smelters outside of ours, and that is sold straight at the alumina index price like Nalco does.

Okay. Wonderful, sir. And last, sir, sir, when you're baking in the IRRs for the aluminium upstream expansion, both at Aditya and possibly at Mahan eventually, given the volatility in LME prices and given the backward integration that the business is undertaking, is it fair to presume that a baseline through cycle of $1,000 per ton is a reasonable number to work with?

$1,000 is a good number to work with, but we get a reasonable return even slightly below that. So the alumina LME projections we take are roughly $2,500 LME when we do those calculations. And remember, though, that $2,500 number is from FY '28 onwards, not today because those projects come in from FY '28. So we are assuming around $2,500 as the base case there.

Okay. Got it. Thank you.

Thank you. Ladies and gentlemen, due to paucity of time, we will take that as our last question for today. If you have any further questions, please get in touch with the Hindalco Investor Relations team. I would now like to hand the conference over to Mr. Pai for closing remarks.

Yes. Thank you. So I think that our integrated business model advantages are seen in a quarter like this where the India business has performed very well. Novelis also on an underlying basis, had a very strong Q2 with an EBITDA per ton above 500. I think that we have quite a lot of confidence that the Bay Minette project when executed is going to be a significant game changer for us in the U.S. market. And I believe that the Indian expansion projects are on track and on budget. So overall, we think that the Hindalco business is quite poised over the next 2 years to become a high-performing business. I just wanted to add, we also believe that once the Bay Minette project is over, Novelis will turn cash positive as well and the net debt-to-EBITDA will start coming down. So all these are on our mind and we are quite conscious that the next year, our execution capabilities will be closely watched by you guys. So thank you very much for your attention.

Thank you. On behalf of Hindalco Industries, that concludes this conference. Thank you for joining us, and you may now disconnect your lines.