Alcoa Corporation (AA) Q2 2017 Earnings Report, Transcript and Summary

Good afternoon, and welcome to the Alcoa Corporation Second Quarter 2017 Earnings Presentation and Conference Call. [Operator Instructions] Please note this event is being recorded. I would now like to turn the conference over to James Dwyer, Vice President of Investor Relations. Please go ahead, sir.

Thank you, Denise, and good day, everyone. I'm joined today by Roy Harvey, Alcoa Corporation President and Chief Executive Officer; and William Oplinger, Executive Vice President and Chief Financial Officer. We will take your questions after comments by Roy and Bill. As a reminder, today's discussion will contain forward-looking statements relating to future events and expectations that are subject to various assumptions and caveats. Factors that may cause the company's actual results to differ materially from these statements are included in today's presentation and in our SEC filings. In addition, we have included some non-GAAP financial measures in this presentation. Reconciliations to the most directly comparable GAAP financial measures can be found in the appendix to today's presentation. Any reference in our discussion today to historical EBITDA means adjusted EBITDA. A note on financial statements methodology. Because Alcoa Corporation commenced operations as a standalone public company on November 1, 2016, the financial statements are a combination of financials carved out from Alcoa Inc. prior to November 1, 2016, and actual results of Alcoa Corporation thereafter. For example, the second quarter of 2016 results are entirely on a carve-out basis, while the first and second quarter 2017 results are entirely actuals. Finally, as previously announced, the earnings release and slide presentation are available on our website. With that, here's Roy.

Thank you, Jim. Welcome to our call to review Alcoa's second quarter results. Overall, our businesses reported solid financial performance in the second quarter even as we experienced lower alumina prices versus the first quarter of this year. We also continued to make progress against our strategic priorities. We grew cash to strengthen the balance sheet and continued to progress on our actions to reduce complexity. We have accomplished much in this first half of the year as a new company, including producing $1 billion in profits, and we remain committed to further improving the business for the benefit of our stockholders. Let's start with a few highlights of the second quarter. We generated adjusted net income of $116 million or $0.62 per share. We reported $483 million in adjusted EBITDA, excluding special items, which is down 9% sequentially as lower alumina prices more than offset higher aluminum prices. Cash generation was strong. Our cash position increased $150 million from the end of the first quarter to $954 million at the end of the second quarter. In our drive to reduce complexity and lower costs, last week we announced plans to restart 3 of the 5 pump lines at our Warrick aluminum smelter in Indiana. It will supply metal to our co-located rolling mill, and improve the overall efficiency of the integrated site. We continue to work to improve results for our canned sheet business and respond to the increased volume that we expect for this rolling mill. Our fundamental outlook for bauxite, alumina and aluminum is relatively unchanged from last quarter. Our alumina balance outlook has improved slightly due to higher alumina demand in China. In the aluminum market, similar to last quarter, recent data indicate that global aluminum demand has strengthened, and we have increased our 2017 global aluminum demand growth forecast to a range between 4.75% and 5.25%. This additional projected demand is offset by increased production in China. China has significant work to do to bring its market back into balance. Although it has already announced and taken some action to eliminate its overcapacity. With this continued strength in both demand and supply, and with 6 months already behind us, we are now anticipating further increases in our input costs. While cost inflation is never welcome, it is paired with improved commodity pricing that drives better profitability. As a result of these changes, we have adjusted our 2017 outlook on net performance to negative $50 million for the year. We have also tightened our outlook on adjusted EBITDA to be between $2.1 billion and $2.2 billion, excluding special items. This is updated from our forecast of $2.1 billion to $2.3 billion in the previous quarter. With that, I'll turn it over to Bill for a closer look at our second quarter financials.

Thanks, Roy. I'll quickly run through the details of the results. And as Jim said, all financials prior to November 2016 are on a carve-out basis. First up is our quarterly income statement. Sequentially, revenues are up 8%, increasing $204 million up to $2.9 billion, with higher third-party shipments across all segments. Compared to last year, revenues are up 23%, increasing $536 million on higher alumina and aluminum prices. Net income attributable to Alcoa Corporation was $75 million or $0.40 a share, declining $150 million sequentially, primarily due to the first quarter $120 million gain from the sale of the Yadkin hydroelectric assets. Let's take a look at the special items in the quarter. Combination of special items this quarter totaled an unfavorable $41 million after tax. The contributors include discrete tax items of $18 million comprised of $28 million unfavorable related to the interim period treatment of operational jurisdictions which no tax benefit was recognized; and $10 million favorable from the difference between the Alcoa consolidated rate and the statutory rate on special items. Other items include restructuring of $11 million, mark-to-market of energy contracts of $6 million and Portland restart power exposure of $6 million. In particular, restructuring items are largely related to labor cost optimization at the Warrick rolling mill and our smelting system. Portland restart power exposure is the near-term market exposure resulting from renegotiating the smelter's long-term power contract, and that exposure ends this July. Now let's look at adjusted EBITDA and the full income statement after special items. Second quarter adjusted earnings were slightly below the first quarter, with net income attributable to Alcoa, excluding special items of $116 million, but improved $160 million year-over-year. Likewise, adjusted diluted earnings per share was slightly lower sequentially at $0.62 per share but up substantially year-over-year. Adjusted EBITDA, excluding special items, is $483 million in the second quarter, down $50 million sequentially but up $173 million year-over-year. Some key points. On a sequential basis, EBITDA margins tightened to 17% due to lower alumina prices. Year-over-year, margins improved 350 basis points on higher alumina and aluminum prices. SG&A and R&D increased $1 million from last quarter due to a slight ramp-up in R&D spending. On a constant-currency basis, however, SG&A decreased versus both comparison periods. For items below the EBITDA line. Depreciation increased $11 million sequentially due to nonrecurring asset disposals in Canada and residue disposal area replacement coming online at the same time previous projects were completing their depreciated lives. Other income expense of $12 million improved $37 million primarily due to foreign currency translation effects. Interest expense is $25 million, consistent sequentially. Remember that the interest expense of $66 million in the second quarter of '16 was interest allocated from our former parent company, Alcoa Inc. Our operational tax rate depends heavily on the jurisdictions where we have profits and losses. In the second quarter, the rate settled at 32.8%. Net earnings attributable to non-controlling interest decreased $8 million sequentially to $72 million due to lower earnings in our AWAC joint venture. Let's take a deeper look at the factors driving adjusted EBITDA. Sequentially, adjusted EBITDA is down $50 million, primarily due to unfavorable alumina prices with partial offsets from higher aluminum prices and other factors. Stronger metal prices were more than offset by the impact of the alumina price index, declining from $342 to $306 per ton, lowering margins $100 million in the Alumina segment. Increased shipments of both alumina and aluminum drove a $15 million gain sequentially. And favorable energy performance from improved pricing in Brazil on lower costs in Spain more than offset unfavorable net productivity in the quarter. The negative net productivity is a combination of smaller items, including temporarily higher operational costs in the bauxite business and slightly higher conversion costs in rolling. Foreign exchange changes in the quarter were a net favorable $17 million. The largest driver of the other category was the planned outage at the Rockdale power facility. Now let's take a look at the business segments. Second quarter adjusted EBITDA in the segments totaled $546 million, down $67 million sequentially, with declines in our bauxite and alumina segments, partially offset by higher aluminum segment earnings. Let's take a closer look at each of the segment's sequential comparison. Bauxite adjusted EBITDA temporarily declined $12 million on lower earnings from our equity partnership mines as well as higher operational costs at Huntly and Juruti. Alumina declined $70 million due to a $98 million impact from lower API prices, with favorable partial offsets from currency, LME and positive net performance. In aluminum, adjusted EBITDA was up $15 million to $221 million. Higher metal prices drove $31 million improvement, along with favorable Brazilian and Spanish power prices. Partial offsets came predominantly from higher alumina costs as well as unfavorable price mix and other factors. As many of you are aware, alumina costs for our aluminum segment are approximately on a 90-day lag. For those tracking particular assets, the Brazilian hydroelectric assets produced adjusted EBITDA of $22 million, up 100% sequentially as energy prices increase in the region. North American flat-rolled aluminum declined sequentially to breakeven adjusted EBITDA in the quarter due to slightly lower productions. Now let's look at adjusted EBITDA items outside of the segments. Combined, our 3 non-segment EBITDA line items improved $17 million sequentially and totaled $63 million. Our transformation entities plus our legacy pension and OPEB costs totaled $40 million in the quarter, increasing $8 million sequentially, primarily due to the planned maintenance outage at Rockdale. In the second quarter of '17, Rockdale was $20 million of the total transformation expense. LIFO and metal price lag, which over time, we expect to net near 0 to turn favorable this quarter and improved $12 million. Other corporate costs also improved $13 million to $26 million, primarily as a result of lower inter-company profit eliminations, mainly in alumina. As we noted last quarter, the eliminations are caused by changes in inter-company profit held in inventory. This quarter, lower alumina segment margins meant there was less inter-company margin to eliminate at the corporate level. If margins and inventories remain stable, future eliminations should be minimal. Next, let's take a closer look at the cash generation. At the end of the second quarter, we had close to $1 billion in cash on our balance sheet, primarily the result of $223 million in free cash flow in the quarter. We also have included a summary table to point out a few key features of our cash flow statement. Typically, and as you can see, in the second quarter, the key cash driver used for -- I'm sorry, the key driver of cash used for financing is dividends to our AWAC joint venture minority partner and the key driver of cash used for investing is capital expenditures. In the first quarter on 2017, the sale of our Yadkin hydroelectric project and the remittance of those proceeds from Arconic as part of the separation agreement impacted both the cash for financing and cash for investing. Now let's move to the other financial metrics. Our balance sheet and related financial metrics continue to strengthen. In addition to our increased cash position, days working capital improved to 18 days in the second quarter compared to 19 days in the first quarter. Capital expenditures now total $159 million for the first half. Second quarter spend was $88 million, up $17 million from the first quarter. Return-seeking capital was $27 million, and sustaining capital was $61 million. We anticipate the spend to accelerate as we move through the year. Our first half annualized return on capital at 6.1% is up 4.6 percentage points compared to our full-year 2016 level. Leverage improvement continues, with net debt of $487 million and net debt-to-adjusted EBITDA of 0.3x. The pension and OPEB net liabilities is $2.9 billion. Let's review our full 2017 outlook. As Roy mentioned, we are narrowing our full-year EBITDA outlook to a range of $2.1 billion to $2.2 billion. This tighter range is based on the same forward price assumptions as last quarter's outlook: LME of 1,900, API of 305, and in Aussie dollar of AUD 0.76 in actual results for the first half of the year. With stronger markets pulling up input costs, we believe net performance will be negative $50 million, with the largest impact being higher raw materials and energy prices. Turning to shipments. Our outlook for all 3 segments is unchanged versus last quarter. For both bauxite and alumina, we expect the run rate in the second half of the year to increase to meet the full-year outlook. On the financial metrics, we're tracking well and have made 3 updates, all favorable. First, we reduced the transformation and legacy pension OPEB spend from approximately $165 million to $150 million. Rockdale outage and power prices have been better than expected, as has been Suriname's energy volume. Second, the cash contributions for pension and OPEB will be approximately $230 million, down from $250 million. Finally, the cash flow impact of environmental and ARO payments has improved $30 million from a range of $150 million to $170 million to a range of $120 million to $140 million. The large reason for the improvement is canceling scheduled demolition work as well as scheduled changes in a variety of close locations. Now let me turn it back to Roy.

Thanks, Bill. To start, I'd like to provide additional details on the rationale behind our decision to restart aluminum production at our Warrick smelter in Indiana and explain why this was the right decision for our business. To recap the details, the restart is expected to be complete in the second quarter of 2018 and includes 3 of 5 pump lines. Restart expenses are estimated between $30 million and $35 million in second half of this year. And lastly, we plan to record an after-tax benefit in the third quarter to reverse our accruals from the previous decision to halt production at the smelter. To be clear, the decision to restart production at our Warrick smelter was not about adding more capacity to the aluminum market, rather we based our decision on 3 things. First, our ability to more fully utilize the assets at the integrated site, which includes the smelter, a rolling mill and power supply. Initially, as you may recall, the assets of Warrick operations would've been divided as part of the company's separation. The rolling mill would've gone to our former parent company. Meanwhile, the power plant and the idle smelter would be with Alcoa Corporation. Now as one integrated location, we can increase the efficiencies at the site. The demolition of the smelter had not yet started and our environmental operating permits remained effective. Second, we evaluated the ability to restart the site with minimal capital expense while directly supplying the mill with molten metal. Essentially, the chosen solution was the optimal use of capital to ensure our Warrick mill was fully supplied with high-quality metal at the lowest possible cost. Third, we also considered that this restart would make our rolling mill more competitive in the canned sheet market through operational efficiencies as the rolling mill prepares to meet expected growth in production volume. Importantly, the decision is also aligned with our strategic priorities. We can reduce complexity by more fully utilizing the assets at an integrated site and will drive returns due to cost synergies and the expected volume growth for our rolling business at Warrick. Turning now to our fundamental outlook for bauxite, alumina and aluminum. We continue to expect our markets to remain stable in 2017. For bauxite, we continue to see the market in relative balance. Our view anticipates that the Chinese will continue to stockpile bauxite strategically throughout this year while in search of new sources of high-quality, consistently delivered material. To maintain our view that China is likely to grow a strategic bauxite stockpile due to refining, we maintain our view that China is likely to grow its strategic bauxite stockpile due to refining expansions in China, offset by production growth in Guinea and the potential export of the remaining port stocks in Malaysia. For alumina, our forecast has somewhat improved with a slight deficit compared to what we anticipated last quarter. We continue to expect that China will curtail refining capacity during the winter heating season for environmental reasons, and we are forecasting slightly stronger global alumina demand driven by additional Chinese smelting production. In aluminum, we continue to see strong demand, and we are increasing our global aluminum demand growth outlook from a range of 4.5% to 5% last quarter to a range of 4.75% to 5.25% this quarter. This revision includes a slightly higher 2.75% to 3.25% demand growth forecast outside of China, driven by strong European demand where we see improved forecast for Rolled Products and extrusion and production. In China, we are maintaining our demand growth figure at 6.5% to 7% on expected strong growth in electrical and machinery applications and in the transportation and construction markets, which are the 2 largest aluminum-consuming sectors. However, this increased demand is offset by higher expected supply in China, so we maintain our general view of a modest aluminum surplus this year. With our aluminum supply forecast still projecting a modest surplus, I'd like to spend a few minutes describing policy activities in China. The Chinese government has made 2 significant announcements about potential supply reforms in aluminum through 2 different government agencies. First, the Chinese Ministry of Environmental Protection announced the reduction of metal production in select cities to improve air quality during the winter heating season. Second, the Chinese National Development and Reform Commission announced that aluminum smelters operating without proper licenses will be curtailed. These actions could potentially impact about 6 million metric tons of capacity per annum. If all of that occurs, we could see a reduction in China's 2017 aluminum surplus. However, only about 800,000 tons per year have been curtailed so far, and our 2017 outlook assumes that these initiatives will have only a moderate impact this year. It is clear, though, that the effects of these policies could be significant, creating risk to both the upside and the downside relative to our forecasts. The potential for meaningful positive change remains, but at present, we maintain our 2017 outlook for a continued surplus of aluminum in China. As we continue to see strength in both aluminum demand and, more importantly, aluminum supply, we have also been experiencing raw material cost inflation over these last critical months. In our aluminum and alumina markets, particularly, we have seen steep increases from last year in the prices for caustic, calcined coke and pitch. As you can see on the right side of this chart, caustic prices have increased by more than 50% compared to the year-ago quarter, pitch prices by more than 25%, and coke prices by more than 10%. We expect these increases to continue, and they are the principal basis for our net performance guidance change. However, this inflationary environment also represents an opportunity for operators, like Alcoa, who are more efficient and with access to high-quality raw materials. Let me illustrate how this helps with a deeper dive into caustic soda's impact. In the last year, we have seen average caustic prices, a key input in alumina production, climb 54% from the second quarter 2016's average of $295 per metric ton to this quarter's average of $455. Alcoa is an efficient consumer of caustic per ton of alumina produced, using less than other ex China refineries, and essentially half as much as Chinese refineries. This difference is created by 2 things: the characteristics of the bauxite supplied by our mines and our process control systems that optimize our refineries. As a result, our refineries see less of a cost increase than most of the market. Higher caustic consumers on the right of the cost curve experience cost inflation faster than those in the left, steepening the cost curve and supporting Alcoa's cost position. On this slide, we quantify the approximate impact from the $160 per metric ton increase in caustic prices experienced over the last year, showing Alcoa refineries cost improving $12 per ton relative to those of Chinese producers and $3 per ton versus the rest of the world average. Increases in input costs are an unfortunate characteristic of a stronger market. However, we believe Alcoa Corporation can create 3 advantages in these situations: number one, strong procurement processes in the advantage of the scale of our purchases blunts the impact versus the general market; number two, the efficiency of our operations strengthens our competitive position; and number three, the attractiveness of our low caustic consumption bauxite is enhanced in our own refineries but also as part of our third-party sales portfolio. As we enter the second half of the year, we remain focused on our 3 strategic priorities which we factor into each of our decisions. We seek to reduce complexity, drive returns and strengthen the balance sheet. To reduce complexity and improve the operational efficiency of the Warrick integrated site, we are restarting some aluminum capacity at the Warrick smelter. It will provide a direct supply of molten metal to the site's rolling mill as it prepares for production growth, which optimizes our use of capital in this business and will also serve to drive returns. We will continue to pursue opportunities to further streamline our business as the year progresses. We delivered solid profitability in the second quarter and generated $1 billion in profits for the first half of the year. As we get a better view of the second half, we have tightened our outlook for 2017. We expect our business to remain strong and to generate $2.1 billion to $2.2 billion in adjusted EBITDA this year, excluding special items. The strength in our markets is creating an increase in input costs, and we've adjusted our net performance outlook for 2017 to negative $50 million. And lastly, we increased our cash position to $954 million last quarter and improved our working capital, both serving to strengthen our balance sheet. Our 3 strategic priorities have guided our actions so far in 2017, and we are making a concerted effort to drive these priorities deep into the organization at every plant and every resource unit by encouraging our employees to pursue all opportunities that can further strengthen Alcoa. With that, I'd like to invite any questions that you might have either for Bill or for me.

[Operator Instructions] And your first question will be from Evan Kurtz of Morgan Stanley.

So I just wanted to dig in a little bit on some of the improvements that you made on the outlook slide and financial metrics to get a sense of kind of how sustainably they are, and how we should be thinking about some of these items in future years. So specifically, maybe to start with the transformation legacy pension OPEB line. You took that down to $150 million. It sounds like a piece of that was Rockdale, which would be sustainable if that's what lowered that. And so how should we think about $150 million as we get into kind of 2018, 2019? Is that a number that we could see going forward? And then similarly, same question really for environmental and ARO payments, that came down as well because you're foregoing the Warrick smelter. It seems like that smelter's going to be around, so should we use these lower numbers in future years?

Yes, Evan. I think you should use those 2 particular numbers for future years. And on the pension and OPEB piece, I believe, in the 10-K, we gave an outlook for future years, so that one will go up in 2018 and 2019. But the 2 you referenced are your best estimate for moving forward. I would comment on the fact that in the transformation expense item, Rockdale does fluctuate up and down. The second quarter was a quarter where we had a planned maintenance outage. That drove some of the expense in the second quarter. And as many of you know, we're working toward a resolution on the Rockdale power contract. So for now, I think those are your 2 best estimates for those numbers going forward.

Great. And for my follow-up, just maybe one on the other corporate expense line. You're tracking way below your full-year outlook. Is that -- is there some chance here that, that could move lower? Or is there going to be some significant second half spending?

That's still the estimate. Yes, we are tracking below but are anticipating a little bit of a run-up in the second half. So we chose not to take that one down to $150 million -- I mean, below $150 million based on our outlook for the second half.

The next question will come from Novid Rassouli of Cowen and Company.

So when you had some EBITDA margin compression on bauxite and alumina, and clearly alumina was on the pricing side, I think, Roy, you'd mentioned about bauxite costs. And I just want to see if you guys could give us some more details on kind of what was driving the cost higher for bauxite?

Yes, this is Bill. I'll give you a couple of items that occurred in the second quarter around bauxite. We had some additional maintenance spending around our fleet in Western Australia. That's temporary maintenance spending that will decline in future quarters. We had a particularly low production at a couple of our joint venture mine sites, specifically MRN and CBG. And in CBG, for instance, there was some labor unrest that caused an outage at CBG. So that has now been resolved and should be okay going into the third quarter. And then lastly, on the bauxite side, a lot of rain in the rainy season that curtailed some production, specifically in Brazil. So those were the 3 big factors around bauxite margin compression in the second quarter. I would point you toward the fact that we haven't moved our full-year production estimates. So we have a fairly aggressive outlook on the third and fourth quarter around bauxite mining production, so -- and we still believe in that.

Great. And then just one follow-up. On the Chinese, I think, surplus that you increased, I just wanted to see, the 800,000 tons that you mentioned that had been curtailed, that is being included in the estimate that you guys provided?

Yes, so let me give you a little bit of background on how we put our numbers together. And if you look at Slide #5, you can see what the total impact could be if everything that the Chinese have talked about had been incorporated. What we've done is gone back and look at the credibility of each of the potential or already announced curtailments and when they could actually happen, and we've prepared what is our best estimate about what will actually occur. So just to give you a range that should be helpful to you, if everything were to happen exactly according to what the official announcements are, it would drive a significant reduction in that surplus, if not drive it into a deficit. However, it's our best estimate of what we think will actually happen in China given past practices, et cetera. So the 800,000 tons, because it's already taken offline is absolutely in the number. But I'll also tell you that part of the 1.7, which has been announced but not yet enacted, we've taken our best, expert opinion on what will actually come down. And we actually have a very talented group in China. We've been working to improve our reconnaissance on what's happening, both from the political side but also from the actual operational side. I would say that over these last few weeks, we've seen a lot of -- not just a lot of rhetoric but actual, some movement to bring down production. So that certainly bodes well, and there is certainly a lot of talk about these actions being real and not just hopes.

The next question will come from Timna Tanners of Bank of America Merrill Lynch.

So about the Warrick restart, I wanted to, first and foremost, ask if there are other smelters that you're considering restarting or assets that you're considering restarting. This is the first such move obviously after a lot of closure, so just wanted a little bit more insight on how you're thinking about your portfolio overall.

Yes, Timna, so let me hit that one. It's a similar response to what we've said in prior quarters. The fact is we look at every single one of our assets and judge and determine whether it's the right time to restart or curtail. So at the moment, Warrick was one that made the most sense. And it's not a -- as I've said in my comments, it's not a need for extra aluminum in the world very specifically because it will strengthen our rolling mill and will be very positive for our shareholders in a number of market -- market configurations. So as we look at places like São Luís, as we look at places like Wenatchee in Pacific Northwest, we also have some curtailed production in Chalco and in Spain, even, each one of those are opportunities for us to find ways to drive return for our shareholders. It requires the right energy pricing. It requires a local demand for value-added products or local demand for even prime P1020. But we look at each one of those and evaluate it based upon the cost structure and ability to drive better returns. So we've made the decision that we've made to date and we'll continue to look at those and see where this takes us. There's obviously a lot of uncertainty right now about what the second half of this year looks like from a Chinese curtailment point of view, and so that's going to also affect how we make those decisions over these next couple of critical quarters.

Like I just said on that, Timna, I mean, Warrick's a pretty unique situation. As many of you know, we have captive, self-generation on one side of the smelter. And on the other side, we've got a rolling mill that we are growing. And one of our key desires is to grow the production out of the rolling mill, and it made a lot of sense under a lot of different -- as Roy said, under a lot of different market configurations to restart Warrick.

And we had talked about in the prior conference calls about Warrick being -- Warrick and our canned sheet business being a bit of a fixer-upper. You should look at the restart of that smelter as an important step towards making that business stronger and better. The fact is we can grow volumes and we need extra metal units in order to do it. And by restarting the smelter, it is the lowest capital cost alternative that gives us the best return. It was a pretty smart decision from my standpoint.

Got you. So just a follow-up, you mentioned uncertainty, and you've got certainly a lot of it with China and a possible tariff action here in the U.S. I just wanted to pivot to ask you if you're talking about canned sheet turning from Warrick, does that preclude you from the strategy of you've been enumerating before about using Ma'aden? Are you thinking differently about that project? And if you can give us any fresh thoughts about what Section 232 may or may not mean for you guys.

Yes, the -- what we're trying to drive from a canned sheet perspective and Warrick, specifically, is separate from what we can do in conjunction with Ma'aden. The fact is, it replaces some of the material that we're getting from the Tennessee that's now a part of Arconic. So a shift into Warrick and driving opportunities there is going to be additive and, in fact, would help to complement the work that we're doing with our Saudi Arabian partners. So we continue to be very positive on our joint venture in the desert. The smelter has been operating very stably now. The refinery has hit better and improved days and we're seeing -- we're coming out of the startup phase and getting into just normal stable production, which is a very good thing to have from a refining perspective. In the rolling mill, we continue to be in the midst of trials and have incredible opportunities to make that a stronger place, so we spend a lot of time with our partners there. As you saw, we were over there as part of the U.S.-Saudi dialogue. We're an important partner inside of the kingdom. And we have, really I would say, a lot of opportunities to make our current joint venture even stronger.

What about 232, Timna?

I thought I've successfully avoided that.

Please. If you would, please. Nice try, though.

Do you want to comment on 232?

Yes, so from a Section 232 standpoint, I think everybody knows the investigation is still ongoing. Alcoa believes that the United States and the U.S. government and Europe and a number of other areas have really been doing a very good job at highlighting the aluminum issue, and that issue very specifically being overcapacity and overproduction inside of China. So just the debate from the 232 case and from the World Trade Organization before that, has created what I think is a very positive set of potential outcomes for aluminum. Now again, we're not going to say that everything China says that they're going to do will be done. However, I would say that the work done by the U.S. government and other actors has really helped to create an atmosphere that has a lot of potential upside for our industry, and then, of course, quite specifically for Alcoa Corporation. That said, I don't think right now there's a determination about which way that 232 case is going to go. I know they're in the midst of discussing steel as we speak, and so we'll react to what the final determination will be.

And ultimately -- and I think you said it, but ultimately, what we want is a level playing field and -- so that's what our efforts have been. And I think, as Roy said, some of the dialogue that has occurred over 232 and WTO has helped that situation.

And it's given us an opportunity, Timna, to talk about what could drive competitiveness inside the United States. And not talking about trade barriers but more talking about how can we get long-term energy contracts that are favorable? How can we look at ways to enhance our ability to smelt efficiently or to creep technology or to find ways to run our smelters in a completely different fashion. So lots of discussions going on. I think it can have a very positive outcome, but still more to come.

The next question will be from Justin Bergner of Gabelli & Company.

Just first question would be to just clarify 2 numbers that sort of came at me a little bit quickly, the Brazilian hydro profit generation and the sustaining CapEx.

Brazilian Hydro was $22 million in the second quarter. Sustaining CapEx in the second quarter was $61 million. Return-seeking capital was $27 million.

Okay. On the bauxite business, can you just talk about sort of what the dynamics in Southeast Asia mean for your ability to grow your Bauxite business to the extent that those dynamics are changing? I think you referred to some of them earlier.

So yes, let me try and hit that and see if it answers your question. Southeast Asia covers a good amount of ground, so let me see if I hit the right countries. First of all, Indonesia. Obviously, Indonesia has just started to allow some exports again. They are relatively modest to date and all indications point to them continuing to be relatively modest. Essentially, there's a requirement that you are operating a refinery or very close to operating a refinery in order to get those export permits. So we think there will be some selective bauxite coming out, but we don't think it's going to reach anywhere near the levels that it had for a period of time. Malaysia is -- it was a similar story before but they continue to ban any kind of production. Now there appears to be some illegal production going on. Those stockpiles are not draining as quickly as one would think given the number of shipments coming out. But the shipments are relatively slim. There really isn't a lot of material coming from Malaysia into China. So stepping back, I think one of the big highlights over the course of this quarter is that Guinea picked up a significant amount of speed over the course of the last number of months, and you now see Guinea as the largest exporter of bauxite into China. Australia continues to be very important. You're also seeing, from a demand perspective inside of China, that you have a number of new announcements about refineries that are going to be bauxite import refineries, which means that they're built on the coast and relatively isolated from the bauxite reserves in-country. I think that's a very good testament to the fact that the reserves inside of China are starting to suffer from a quality perspective and are starting to increase in cost because of the additional need for stripping or infrastructure or even going underground in some cases. And so it continues to mean that there is very healthy demand of bauxite. I also believe -- and if you just step back to the caustic story that I spent a little time talking about it, I happen to believe that the bauxite that we can offer, whether it's coming from our joint ventures in Guinea or in Brazil or coming from our Western Australian mine, is an excellent opportunity for us to grow our value but also to have a great product that can be imported into China. Now we connect that over with the broader questions about what is China's production portfolio going to look like? One of the great opportunities we have at Alcoa is that we can choose to allocate our capital into the part of the business that's the most attractive. I think, for us, bauxite has real opportunities for us, and we have customers that are essentially lining up in order to take additional production as soon as we can get it out of the ground and get the right infrastructure in place. But we also have opportunities in alumina and aluminum, and so we will continue to pretty thoughtfully look at each and every one of those places to look for growth. Did I hit the right points of your questions or did I miss something?

Yes, no, that was extraordinarily thorough. Just one quick question, and then I'll get back on queue. On the Warrick restart, how does this affect your ability? Are you actually going to be selling -- have the opportunity to sell power onto the open market under this restart? Or are you just going to be consuming more power in -- have the opportunity consume more the power in the use of your own assets?

The vast majority of the power that we will generate is going to be used in the restart. So just to give you a little bit of perspective. We've got joint ownership of a fairly large unit there, a joint ownership with Vector, and we've got 3 smaller units. Those 3 smaller units were completely being sub-optimized in a curtailed state. So the economics for this restart are going to be seen really within 3 areas: one is better utilization of the power plant, two is running the smelter and making earnings in the smelter, and then thirdly is the growth and opportunities that we have in the rolling mill. So again, it's largely a no-regrets move at this point.

Great. And each of those 3 areas will be earnings accretive individually?

Each of those 3 areas will be earnings accretive into the aluminum sector, so you won't get a lot of transparency into it other than to see earnings grow when the facility is up and running. And as we said, first hot metal in the first part of next year, full operation in the second quarter.

And the next question will come from Dave Gagliano of BMO Capital Markets.

It's just a quick one. What was the average alumina cost that flowed through the primary aluminum segment in the second quarter?

You know what, Dave, off the top of my head. I don't know, and you can follow-up with Jim after the call.

The next question will be from Jorge Beristain of Deutsche Bank.

On your growing free cash hoard, just wondering if you could again reprioritize for us what your uses of that cash are over the next 12 to 18 months, and what your view is toward re-establishing a dividend policy.

Yes, sure. So good question, Jorge. The -- it was very good cash flow generation in the second quarter, $223 million free cash flow. I'd like to see that the cash balance grew. Had $954 million of cash in the bank at the end of the second quarter. Some of the priorities that we would have for that cash, first and foremost, we're looking to continue to delever. We think that the rating agencies probably have our credit rating incorrect now because when you look at our EBITDA generation and the fact that we're targeting $2.1 billion of EBITDA this year, we think that there's -- that they fundamentally have the rating wrong now. But in order to show them that they have the rating wrong, we're most likely to continue to pay down a little bit of debt. We have pre-payable debt down in Brazil of around $200 million. Secondly, you know we have increased the amount of return-seeking capital that we are looking to spend. We did that at the very beginning of the year so that's not new news. We're looking to spend $150 million of return-seeking capital this year. And then beyond that, first of all, we will hold cash on the balance sheet. We would like to hold $1 billion of cash on the balance sheet. But once we get north of $1 billion, we will be looking at opportunities to either invest in the business or potentially pay into the pension plan to do further delevering. As far as returns to shareholders, as many of you know, our revolver today is fairly restrictive on our ability to return cash to shareholders. We would like to get a better credit rating over the next few quarters in order for us to go back and renegotiate those terms on the revolver so that we can begin to talk about returning cash to shareholders either through stock buybacks or dividends. But that's where we stand today, Jorge.

Got it. And just if I could have a follow-up as well. I think back in 2Q '16, you had mentioned that the closure of Warrick was going to be about a $50 per ton impact. Would we expect to see a similar $50 per ton benefit once it's fully operational?

We haven't provided what the dollar-per-ton impact is. I think if you were to go back and look at the announcements that we made around the first couple of quarters last year to see the impact on cold metal, we will reverse that impact once it's up and running. So I don't have a number that I can give you, but we had provided in prior quarters what the negative impact of cold metal was. And quite honestly, I just don't have it off the top of my head, Jorge, but it's there.

The next question will be a follow-up from Evan Kurtz of Morgan Stanley.

Just a quick one on Midwest premiums. Just want to get your thoughts on the recent weakness. I heard an interesting story about maybe some sub-optimally shaped P1020 in the market that was imported and people were willing to pay less for it. Just hoping if you could confirm or dismiss that and kind of what your thoughts are on that move in the market.

Evan, I appreciate the question. I had not heard the sort of oddly shaped P1020. When I took a look at it and we spent a little bit of time discussing, particularly because of the most recent decline, to me, it sort of steps back to what's happening from a contango standpoint. We have seen contango shrink pretty significantly over the last couple -- for the last few months. And when that contango shrink, it typically means that you're starting to see metal delivered and actually coming out of warehouses. And the lowest cost warrant that's out there right now sits over in Japan or in Korea, really, in East Asia. And so that material started flowing to what was a very short market which tends to be North America. And you started to see more material show up in North America, and maybe that's some of these sort of oddly shaped P1020 that's coming over. And so I think what you are seeing is that with contangos start to reassert themselves and it's actually come back to a much healthier position last I looked at it. I think you're going to start to see some of that material continue to be financed or start building up some more of these financed stocks, and we'd like to see that premium, of course, go back up to prior levels. So it really just seems to be more of a movement of metal from one part of the globe to another part of the globe and really isn't an indicator of the underlying demand. I said that North America continues to be a very strong demand. You're seeing an even stronger demand in Europe than we've seen the last time we've talked about this, and so it doesn't signal to me that there's problem -- underlying problems in the markets themselves.

[Operator Instructions] And we have a question from John Tumazos of John Tumazos Very Independent Research.

If I could follow-up on your balanced or almost balanced alumina outlook to the world market. In the February to May months for International Aluminum Institute data, the alumina output averaged 6% more than the ingot output required or about 2.3 million tons too much, and this is just naïvely using the IAI output. In your balanced outlook, do you expect the alumina output to go down because of the Chinese environmental restrictions next winter? Do you think the ingot output is more? Obviously, the alumina is hard to store, but at least the first half of the year doesn't seem to be balanced.

Yes, John. I think I know where you're going with the question, and let me just walk you through a couple of points. I think coming into this year, we'd already cleared out a lot of the inventory that were sitting out there. You look at the end of 2015, beginning of 2016 when prices were so depressed there, we had built up about as much inventory as is possible inside of the alumina market and we've worked that off over the course of 2016. So coming into this year, I think that was a pretty balanced market between buyers and sellers. And so as we progressed through it, the way that we look at both production and demand, we found it to be pretty equally balanced. And we look at it on a quarter-by-quarter basis. And in fact, when we look at the first quarter, it really looked to be a slight deficit. Now the fact is, like you said, you cannot -- it's not particularly easy to store alumina, and what we've seen is that we've got pretty quick reactions in the market so that we actually, because of the spot pricing mechanism and the load -- or vessel-by-vessel characteristics of the way that this is sold, you can really see what's happening on almost a day-by-day basis in this market. And up until now, we've not seen an excess of tons either sitting in China or sitting outside of China, which is what's driven a lot of strength in pricing. I think you started to see here over the course of the second quarter, prices have come down some. You had more sellers than you had buyers, but we've essentially rectified that situation. We're back to a point where it's relatively quiet from a sales perspective. We continue to see that there's just not a lot of alumina out there that can be purchased and, therefore, not a lot of opportunity for downside. So I'm not as familiar with the IAI data. We have a very clear look at how we do this ourselves and we watch what's actually being produced in China. From our perspective, our balances are the ones that make most sense to us and it's the ones that best describe what's happening from a pricing and a sales perspective on each of those vessels.

And ladies and gentlemen, this will conclude our question-and-answer session. I would like to hand the conference back over to Roy Harvey for his closing remarks.

Yes, in closing, I just wanted to say that we've had solid results in this first half of the year. We have a clear path in front of us for the remainder of 2017. There's a lot going on, whether you talk about China, if you talk about the aluminum industry globally. I think there are great opportunities, but we also are prepared no matter what comes at us and prepared to respond as might be necessary. We're going to continue to find ways to reduce complexity. We're going to search for ideas like Warrick that can drive better returns for our stockholders. We're going to strengthen our balance sheet. We continue to put cash in the bank, and we look forward for continuing to describe that story to all of our investors in the coming quarters. So thank you very much for your attention. I truly do appreciate it. And I turn it back over to our operator, Denise, now. Thank you.

Thank you, sir. Ladies and gentlemen, the conference has now concluded. Thank you for attending today's presentation. At this time, you may disconnect your lines.