Ferroglobe PLC (GSM) 2018 Q2 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the Ferroglobe Second Quarter 2018 Earnings Investor Conference Call.

(Operator Instructions) As a reminder, this conference call may be recorded.

I would now like to introduce your host for today's conference, Mr. Phillip Murnane, Chief Financial Officer.

Sir, you may begin.

Good morning, and thank you for joining the Ferroglobe Second Quarter 2018 Conference Call.

I'm going to read a brief statement and then hand the call over to Javier López Madrid, Ferroglobe's Executive Chairman.

Statements made by management during this conference call that are forward-looking are based on current expectations.

Risk factors that could cause actual results to differ materially from these forward-looking statements can be found in Ferroglobe's most recent SEC filings and the exhibits to those filings, which are available on our web page, www.ferroglobe.com.

In addition, this discussion includes references to EBITDA, adjusted EBITDA and adjusted diluted earnings per share, which are non-IFRS measures.

Reconciliations of these non-IFRS measures may be found in our most recent SEC filings.

Slide 2, please.

On the call today, we will review the Q2 results across our core products, followed by some consolidated financial highlights.

Lastly, we will provide an update on the company's operating environment and highlight some areas that have significant potential for long-term growth and value creation.

At this time, I will turn the call over to Javier López Madrid, Ferroglobe's Executive Chairman, to provide a few comments.

Next slide, please.

Good morning, everyone, and thank you for joining today's call, concurrent with our second quarter earnings release to announce the Board of Directors' approval for a share repurchase program.

This unanimous decision by Ferroglobe's board follows a related vote at our recent General Shareholders' Meeting whereby 99.9% of shareholders voted in favor of the share repurchase plan.

The board's decision to authorize the share repurchase program is based on our belief that Ferroglobe's recent share price level do not reflect the company's long-term intrinsic value and presents strong investment opportunity for the company.

Let me take a moment to touch on several reasons why we believe that these price levels do not reflect Ferroglobe's real value.

One, the last time our shares traded around this level was in September 2016, a period when we were generating significantly lower EBITDA and we were dealing with significantly higher leverage.

To put things into perspective, Ferroglobe generated approximately $176 million of adjusted EBITDA in the first half of 2018 alone.

This is nearly as much as what the company generated in all of 2017, and 2017 was itself a year in which we have more than doubled EBITDA over the previous one.

Moreover, Ferroglobe's [net] leverage was -- which was above 5x at the end of Q2 2016 is currently well under 2x, further highlighting the dramatic improvement in our financial well-being in the last 2 years.

Today, our end market across steel, aluminum and chemicals sectors are all performing well with demand expected to remain strong.

This is quite different from 2016.

Number two, operationally, our company has competitive advantage in its ability to service customers globally.

There is no other producer in the world, which offers a combination of sites, geographic presence and low cost of production.

The resulting operation of flexibility puts us in an unrivaled competitive position.

This advantage is particularly important, given the current backdrop and noise relating to trade wars and tariffs, particularly in the steel and aluminum sector.

In particular, Ferroglobe is in a prime position to continue to serve 2 protected markets, the U.S. and Europe, given our production presence and capacities in both regions.

Number three, shifting back to our recent performance, we're happy with our results in the first half of 2018.

However, we're far from realizing the full potential of our platform.

The manganese alloy business is one area where we expect significant upside.

With the recent acquisition of Tier 1 manganese alloys facility, we have doubled our capacity in this product category.

However, we generated slightly lower adjusted EBITDA compared to the previous quarter and less than a normalized return for the first half.

Pedro and Phil will go into specific reasons for this, but I would stress that we have a good quarter overall despite the manganese alloys part of the business contributing relatively little.

During the quarter, we deployed funds towards working capital in these plants and remain confident this portion of our business will be a stronger contributor going forward.

Lastly, the board consider longer-term value creation drivers including organic growth, M&A, leveraging of our technology and divestitures of noncore assets.

In particularly, we recent highlighted a very exciting milestone achieved during testing of photovoltaic cell made using our solar-grade silicon.

We're pursuing similar projects tied to lithium ion batteries and solar more broadly.

While it's too early to determine the full potential of this project, our success to date give us confidence that we can create value in these product areas, and we intend to pursue the most promising of these project forward.

For all the foregoing reasons as well as others considered by the board in approving the share buyback, we collectively feel Ferroglobe has the potential to generate stronger EBITDA in the future and believe our share price will come to reflect this upside.

With our result, one of the prudent uses of our capital to date is to repurchase Ferroglobe's shares at current valuations.

And while the board support the share repurchase program, we're equally committed to striking an optimal balance in a broader capital allocation strategy including further deleveraging and consideration of all the value-enhancing opportunities as they arise.

In closing, may I say we're extremely proud of the progress we have made to date and look forward to delivering value creation to our stakeholders?

With that, I would like to turn the call over to Pedro Larrea, Ferroglobe's Chief Executive Officer.

Thank you, Javier, and good morning, everyone.

Next slide, please.

In Q2, Ferroglobe performed, as expected, confirming the positive fundamentals of the business.

We posted a quarterly net profit of $66 million and an adjusted net profit of $25.7 million in Q2 compared to $33.3 million in the previous quarter.

Reported EBITDA in the quarter was $130.9 million.

On an adjusted basis, EBITDA in Q2 was $86.3 million, down 3.7% from $89.6 million during the prior quarter.

Overall, the differences between the reported and adjusted figure is derived from the bargain purchase price gain that has been recorded relating to the manganese alloys plant acquired earlier this year.

In Q2, we achieved top line growth of 4% over the prior quarter.

However, higher input costs resulted in adjusted EBITDA margin of 14.8%, a decline of 120 basis points.

Next slide, please, so on Slide 6. In the first half of 2018, we have seen a step change in our financial performance that brings the company back to more normal operating environment.

During the first half of 2018, we generated $175.9 million of adjusted EBITDA, reflecting 135% increase over the same period last year.

The adjusted EBITDA generated in the first half of 2018 amounts to 95% of the company's total adjusted EBITDA in 2017, once again highlighting the return of our business to a more stable state.

While the year-over-year trend is certainly positive, we believe the full potential of the business is not reflected in this comparison.

The continued cost pressures, along with lagging manganese alloys' performance, impacted our results in the first half of the year, particularly in Q2.

Once these pressures start to ease, we stand to benefit meaningfully, including through additional contributions from the manganese alloys business.

Overall, Ferroglobe's year-over-year performance clearly illustrates a positive dynamic around growth in the underlying business, in turn, supported by end-user growth.

Next slide, please.

Our consolidated sales grew by 4% in Q2 over the prior quarter, continuing the steady increases in our top line we have seen in the past few years.

For the quarter, we have strongest revenue growth in our manganese alloys business due to the addition of the 2 acquired plants.

Ferrosilicon sales were flat quarter-over-quarter while silicon metal revenues showed a slight decline due to lower overall activity in the industry, which we believe was a result of inventory builds at some of our customers in anticipation of the trade case outcome.

The ability of our product contribution to change its quarter reinforces the value of our diversified product portfolio.

Year-to-date, we have significantly increased volumes, and we have significantly improved our average realized selling prices across the portfolio compared to 2017.

The high utilization rates and strong demand provides a good dynamic to support this trend line for the business.

Next slide, please.

Adjusted EBITDA declined 3.7% over the previous quarter with stronger volumes and relatively stable prices in Q2 being offset by an increase in raw materials' cost and extraordinary maintenance.

Overall, the quarterly adjusted EBITDA was positively impacted by incremental volumes in manganese alloys and silicon-based alloys.

In the aggregate, these factors contributed an incremental $3 million in adjusted EBITDA quarter-over-quarter.

Average selling prices across our core products were marginally lower in those 2 product categories while realized silicon metal prices were flat.

In the aggregate, the realized selling price evolution across all products resulted in a negative adjusted EBITDA impact of $1 million during the quarter.

The biggest contributor to the decreasing quarter-over-quarter adjusted EBITDA was cost pressures, having an aggregate negative impact of $8.9 million.

Our industry is faced with ongoing inflationary cost pressures across our products, resulting in an adverse impact of approximately $7.5 million quarter-over-quarter.

Manganese ore prices, energy prices and electrode cost increases all contributed to this impact.

Additionally, several facilities were faced with technical issues and disruptions during the quarter, resulting in $4 million of additional spend.

On the positive side, our overhead costs at headquarter's offices were reduced by $2.5 million.

Foreign exchange volatility, driven by a strengthening U.S. dollar, had an adverse impact of $1.8 million in Q2.

A $6.2 million improvement in the other category include the negative impact of $4 million in the energy business, which is still at exceptionally high EBITDA levels of $5.6 million in Q2.

It also includes positive contributions from noncore products and from mining.

Next slide, please.

On the next slide, we will discuss pricing and volume trends, earning contributions and market observations for each of our key product families.

Turning first to silicon metal on Slide 9. Despite some price -- pricing declines in the U.S. and in the European indices, Ferroglobe maintained a flat realized average selling price for silicon metal, reflecting a well-managed commercial strategy and a good mix of fixed and index price contracts.

Our average selling price increased by 0.4% to $2,773 per metric ton.

Prices across North America, Europe and China remained well above 2017 levels during the same period are now at attractive levels following the trade case rulings, underpinning the robustness in end market demand.

In Europe, the decline in the index pricing is largely attributable to seasonal lower activity.

U.S. pricing has remained at above $1.3 per ton, significantly higher than expectations after the negative outcome of the U.S. trade case.

Overall, the fundamentals of the silicon metal industry remain robust.

The silicon, solar and the aluminum industries are all faced with tightness in their respective sectors.

The bar chart on the top right of Slide 9 shows the first quarter-over-quarter decline in volumes over the past year.

This was primarily attributable to lower sales activity as customers built up stocks in advance of the trade case outcome.

The cost increases we face in silicon metal production continue to be largely attributable to electrodes and energy.

We were also faced with some extraordinary maintenance-related costs at the Selma, Polokwane and Montricher facilities, which adversely impacted silicon metal EBITDA by $3 million during the quarter.

Furthermore, we continue to monitor the economics of the silicon metal business relatively -- relative to ferrosilicon to evaluate if any further capacity conversions should be implemented.

Next slide, please.

Turning to silicon-based alloys.

Overall, EBITDA contribution from this area declined due to slightly lower pricing and increased costs.

During the quarter, the average selling price decreased by 2.4% to $1,908 per metric ton.

Despite this drop, the pricing of ferrosilicon, which accounts for the majority of our silicon-based alloy sales, remains near historically high levels.

Sales volume continues its trend of steady quarter-over-quarter increases, reaching 78,214 metric tons in Q2.

We believe the relevant demand drivers will continue to support ferrosilicon prices at these levels.

Globally, the steel market has been robust despite the recent trade war-type actions.

In particular, Ferroglobe is in a prime position to continue to serve 2 protected markets.

The United States implemented Section 232 and Europe enacted safeguard methods, fostering local steel production and increasing demand for our products.

Additionally, the ongoing environmental crackdown in China has resulted in ferrosilicon production capacity curtailments.

Similar to silicon metal, the silicon-based alloys business was faced with increased costs in the second quarter.

Some of this was due to annual overhauls, such as at our Bridgeport facility, while we also face higher energy costs in Europe.

I would like to stress that around 1/3 of our silicon-based alloys business is in foundry products.

These are tailor-made, noncommodity products [with much weighted] stability in prices and where -- and where we are making significant progress with 10% growth year-over-year.

Next slide, please, so turning now to manganese-based alloys.

Q2 represented the first 2 quarter with the recently acquired manganese alloys facilities in France and Norway.

As a result, we saw significant increase in sales volumes during the quarter.

Despite this increase in volumes, the quarter EBITDA contribution from this product fell to $6.6 million, down 45% from $12.0 million the previous quarter.

The manganese business has been under pressure as the prices of these alloys in Europe have dropped in recent months while manganese ore prices remain high, all having an adverse impact on our spread.

The average selling price for manganese alloys decreased during the second quarter by 5.2% to $1,304 per metric ton.

Product margins also were affected by increasing manganese ore prices and high -- higher energy costs in Spain.

Improvement in manganese ore prices will benefit us in late 2018 and early 2019, and energy prices in Spain should eventually return to more normalized levels.

And as we have mentioned for silicon-based alloys, Ferroglobe is in a prime position to continue to serve 2 protected markets, where protective measures are fostering local steel production and increasing demand for our products.

With that, I would now like to turn the call back to Phil, who will review the financials in some more detail.

Thank you, Pedro.

If we can move to Slide 13, please.

As you've already heard, sales volumes are up 13.6% to over 271,000 tons in the quarter and revenues are up to approximately $583 million.

The quarter saw growth in our manganese alloys business volumes due to the addition of 2 acquired plants and ferrosilicon sales were flat whilst silicon metal revenues showed a slight decline.

Adjusted EBITDA of $86.3 million compares favorably with prior quarter and represents a 14.6% EBITDA margin, which, although down from the prior quarter, was still well above the same period in 2017.

Moving to Slide 14.

Overall, working capital increased to $407 million.

This quarterly increase includes $35 million in further working capital builds in the 2 acquired manganese alloys plants.

The growth in working capital in both the quarter and the first half has had a corresponding impact on both net debt and free cash flow.

We have implemented initiatives to return this working capital during the second half, which I will discuss in more detail.

Although net debt has increased in the period, our balance sheet metrics remained strong, and we will continue to focus on deleveraging.

We are committed to this target to leverage that we announced in Q1.

If we move to the next slide.

As I noted previously, the working capital increase in the second quarter includes a $35 million increase in working capital, you see a total use of working capital of $90 million in the acquired manganese plants.

The Q2 impact was primarily driven by normalization of the payment terms on manganese ore.

Additionally, you see working capital increase in the second quarter for an increase in finished goods inventories across various products.

I would also highlight that the increase in our AR securitization program has started to capture additional sales volumes.

Q3 should see a further increase as the securitization program has been extended by an additional $50 million.

Next slide, please.

As I noted on the previous slide, the working capital increase in the second quarter for the manganese -- included $35 million increase in working capital in the acquired manganese plants to a total of $90 million, primarily driven by normalization of working capital (inaudible).

I will try to extend -- going into more detail on the acquisition of the 2 manganese plants.

The $90 million of working capital balance at the end of June includes approximately $98 million of inventories, of which $43 million belongs to raw materials and $55 million to finished products inventories.

These inventory levels represent 155 days outstanding versus an average of 64 days outstanding for the rest of the company.

Even allowing for additional logistics complexity, normalized working capital in these assets should stand at $60 million in due course.

Accordingly, we expect a $20 million reduction in inventory levels by year-end.

During Q2, we recognized a bargain purchase gain for the acquisition of $46 million, representing the fair value of net assets from the acquisition, less consideration [of noncash] and contingent consideration.

Fair values for property, plant and equipment assets recognized were based on an independent valuation with an immaterial change from recorded book value whilst working capital and other assets were booked without any adjustments.

Fair value with the contingent consideration for the earnout liability was based on an NPV calculation, including historic spreads and their frequency, the earnout percentages and [discoveries].

No gain or loss was recognized in Q1 2018 as the purchase accounting have yet to be completed.

If we move to the next slide.

As highlighted previously, working capital investments from the new plants and assets were $90 million.

As you can see, this has driven an increase in our overall gross and net levels.

If we move to the next slide.

Looking in more detail at our cash generation.

Profit for the 6-month period was $102 million.

Adjusting for noncash items, including the bargain purchase gain, our total adjusted for these noncash items was approximately $178 million.

This cash was then invested in operating assets and liabilities, $158 million, [which is] a significant working capital investment that you saw in the 6 months.

Interest payments of $20 million, income tax payments of $24 million and payments due to investments of $55 million -- $52 million, sorry, resulting in a free cash flow from the first half of $77 million.

If I highlight a few lines.

The working capital increase during the period includes the newly acquired plants, as I have noted.

The scale of this investment would not occur in a normalized period.

Regarding the interest payments of $20 million, we are actively considering refinancing the company's 9.375% 2022 senior notes, and we expect any new debt to reduce interest expense.

And finally, in respect to payments due to investments of $52 million, we would expect for current CapEx going forward to be below H1 2018 levels.

Given the outflows required during H1 2018, we have implemented various cash-generating initiatives, including a planned increase in the AR securitization program, will provide additional liquidity of $35 million.

As previously highlighted, we will reduce manganese ore inventory levels with a target reduction of $20 million.

We are planning on increasing the rotation of finished product inventories in key products and adapting production capacities for commercial commitments, which we expect to deliver $20 million.

And finally, we expect to complete noncore asset divestitures during the remainder of the year of about $20 million.

We expect to be cash flow positive in the second half of 2018, and we are actively committed to targeting positive free cash flow for the full year.

Next slide, please.

With a strong Q2 and H1 in 2018, we continue to be committed to a conservative capital structure, and we'll focus on continued deleveraging of the company's balance sheet.

We will continue to balance this commitment with returning value to shareholders and consideration of other investment opportunities.

Now I'd like to hand back to Pedro, who will comment on end market dynamics and provide an update on growth drivers.

Thank you, Phil.

So if we turn to Slide 21.

To provide some further perspective on the back half of the year, we have included a breakdown of our order book across our products.

In silicon metal, we had approximately 84% of our order book contracted for the second half as of the close of Q2.

The fixed price contracts in North America and Europe are above the current index price.

Some index contracts are still reflecting the positive evolution of the indexes during the first half of the year, so we expect overall index contracts to remain at stable price levels in the coming months.

And we feel the overall supply/demand tension in the market as well as increasing input costs provide good reason to expect prices to remain roughly stable around these levels.

In silicon-based alloys, we had approximately 87% of the order book contracted for the second half as of the close of Q2.

Not only have we locked in fixed price contracts at very attractive levels, we have also optimized our production capabilities to maximize the margins in this business.

Increased demand in our markets and plant closures in China are both factors that helped the overall tightness in this market.

Lastly, in regard to manganese-based alloys, we are expecting improvement in this business as the alloys' prices have not increased in line with the ore prices in past quarters.

Going forward, we expect the spread between the alloys' price and ore cost to get back to historical levels.

Hence, the order book for this product category is weighted more towards booked business and the index and spot volumes going into the end of the year.

This sentiment ties into our commercial strategy, which is to remain disciplined during periods of sound fundamentals.

Should these dynamics change in the future, we maintain the flexibility to use our global production platform to our advantage and can take advantage to optimize our performance accordingly.

We believe current price levels in all our product markets and, specifically, in silicon metal and in manganese alloys, are resulting in tight margins for most producers worldwide.

So we see no reason for suppliers to reduce their offer prices and risk negative profitability.

Next slide.

It is important to look at the medium and long-term outlook of -- for our markets.

Ferroglobe's products are sold into a diversified product by a group of end markets, which are not generally correlated to one another.

Given the overall health of the global economy, our core end markets have all performed well and remain strong.

A recent [wave of] earnings updates and outlooks by many of our major customers across these verticals reinforces our confidence that demand in the next 18 months should remain healthy.

We do not think the evolving trade wars and tariffs, which have targeted the steel and aluminum sectors, in particular, will impact the aggregate demand for steel or aluminum globally.

What could change is the location of where the aluminum or steel is sourced.

We believe that the production of these products in our geographical markets, North America and Europe, will remain strong and, hence, the need for Ferroglobe's products.

Another factor potentially impacting our markets is the Chinese environmental regulatory shutdowns and financial reforms, which have impacted the steel, aluminum and chemical industries as well as the production of silicon metal, ferrosilicon, manganese alloys, resulting in lower Chinese export trends.

With regard to chemicals and silicons, we perceive favorable sentiment among our customers who recently released earnings and expressed view of their end markets going into the back half of the year.

The tightness in these markets continued, and incremental supply will probably not catch up to meet demand until 2019.

Turning to polysilicon.

The outlook beyond 2018 remains positive as the overall growth in this end market will continue as the rest of the world, ex China, continues to see an increase in demand for solar panel.

Next slide, please.

The images on the left side of Page 14 (sic) [Slide 23] show the construction of our net -- new joint venture facility in Puertollano, Spain as of a few days ago.

We remain on track to complete construction in the coming months.

Further, there have been no changes to the budget we initially presented related to the consumption costs.

On August 14, we issued a press release detailing an important milestone in our continued development of solar-grade silicon capacity.

Over the past few months, our material was successfully tested by one of the global leaders in the photovoltaic industry after it manufactured over 130,000 solar wafers using 100% unblended solar-grade silicon produced by us.

The results obtained and certified by a well-known independent German testing institution confirmed that our solar-grade product compares favorably with chemically produced polysilicon that is manufactured using a process that is orders of magnitude less energy-intensive and entails much lower environmental impact.

These are exciting validations of our proprietary processes and products, not only for ourselves but for our potential customers whom we are working with.

Indeed, we are now commencing discussions with such potential off-takers.

Next slide, please.

The long-term growth of this business will come in part from the various projects we are currently developing.

We are investing in R&D of products across several verticals by leveraging our silicon metal and manganese alloys expertise.

This includes in addition to production of solar-grade silicon for the photovoltaic industry, high-value powders for industrial use and the development of next-generation lithium ion batteries that include our products in both the cathode and the anode.

These projects are currently at different stages of development.

As we hit certain milestones, we will consider collaboration with strategic partners to help finance the growth and accelerate our plan to market on a commercial scale.

Our industry has always been characterized by innovation, and we feel it is critical for our long-term growth to stay ahead of the curve.

Developing these types of projects also helps us shift our mix away from commodity-grade products, which we expect will ultimately improve our profitability.

Overall, we anticipate that the combination of these efforts should add 15% to 20% to Ferroglobe's top line results in the near to medium term.

In closing, the second quarter confirms a solid first half of 2018 and continues to support a positive evolution of our business.

The general sentiment we have shared today is one of confidence in our ability to leverage our unique platform and capabilities to maximize the potential in the second half of the year.

While demand from our end markets' customers remains solid, we need to counter cost pressures to maximize our margins and stay attentive to pricing trends in our product markets in order to act flexibly in the deployment of our production capacity.

We stand by our decisions in terms of converting furnaces and making investments in the manganese alloy side of the business.

These strategic moves are important to the long-term success of the company as they provide increased flexibility.

By focusing on continuous improvement and operational excellence, we commit to improving our balance sheet further by year-end.

Furthermore, by staying disciplined with our commercial strategy and strategic plan, we will return value to shareholders.

At this time, I would like to open the line to any questions.

(Operator Instructions) Our first question comes from Ian Zaffino with Oppenheimer.

Thanks for the color regarding like the pre-buys on the silicon metal side.

I may have missed this, but can you maybe touch on what the inventory levels you're seeing at your customers are for silicon metal?

How much were they actually buying ahead of the trade cases?

And where are we sort of in the work done of that inventory before, I guess, you could start realizing like normalized sell-through to those customers?

Thank you, Ian.

We don't have a lot of transparency, I must say, in inventories at our customers.

So it is not obvious, I would say, what is the exact numbers of inventories at our customers.

What I would say is that we have felt during Q2 some slowdown in consumption from some of those customers, and our hypothesis is that they have been accumulating some inventories in Q1.

That should have been, I would say, depleted during Q2.

Also, it is important to note that significant customers in North America were idle because of industrial incidents in their facilities, and those are back up.

So we see that, again, our customers are running pretty much at full capacity and that, again, inventory should be more or less today at normal levels.

Okay.

And also, I know you've been a big believer in the recent price decline in silicon metal being driven by the availability of hydro in China.

When does this kind of period, call it end, and then we'll start to see maybe supply coming out of China coming down and prices in China going up and kind of the rest of the world then going up as well?

What sort of timing of that should we expect?

Well, again, I would say 2 things.

One is, yes, there is typically a bit of a seasonal dip in prices in China this time of the year.

It has 2 sources.

One is, as you say, lower power prices in some regions of China.

The other is actually lower activity in China because of holidays or vacations, and of course, those 2 factors are seasonal.

What is important to stress is that such seasonality this year has happened at prices that are $300 to $400 per ton and higher than a year ago.

So that, again, just shows a very healthy trend.

The other point I have been stressing in the past few weeks is actually cost inflation in China.

The way we see evolution of the silicon industry in China is one with much higher costs, which provides, in our view, a good support for prices going forward.

So our view of evolution of prices in China and in the rest of the world is, again, that there is a solid support at more or less current levels of silicon metal and an upside as soon as those seasonal events fade away.

Our next question comes from Martin Englert with Jefferies.

You've seen a number of sequential cost pressures in 2Q, about $9 million quarter-on-quarter.

How do you view that incrementally going into 3Q?

Well, as you see, part of those cost increments are exceptional and have to do with -- again, with exceptional or extraordinary maintenance in some of our facilities for a number of around $4 million, I would say, our extraordinary costs in Q2.

If we look at overall raw materials' cost and the rest of the cost components, right now we see stable trend in costs.

And eventually, some of the factors that have increased cost in Q2, like energy in Spain, as we were saying, eventually, should be going down.

However, we are all, I think, aware of the volatility in many factors, and mainly in commodities, so coal, manganese ore and energy worldwide, again, we think, are stable at current levels, but they are volatile.

Electrodes' costs, which has been a source of a lot of debate in the past few quarters, are again stabilizing.

We need to also underline that we started our electrode production facility in China and that is providing some hedge to those increased costs in electrodes, and we are there to benefit.

We can extend and expand the reason why we believe that the electrode situation worldwide is actually a positive for us as Ferroglobe.

But as I was saying, we do have some degree of hedge with the production in our Yonvey facility.

Okay.

So net-net, [you're already] seeing stability in some of the variable costs here and then the maintenance items will likely roll off quarter-on-quarter.

Have you considered or are you planning to idle any facilities due to the continued deterioration in silicon metal and alloy prices when we look at the back half of the year?

Well, I would -- [I don't see] continued deterioration of silicon metal prices.

So I would say those -- that prices are -- in the past few weeks, have stabilized.

As I was saying, we believe that there is a solid ground.

We don't see any reasons why suppliers worldwide would be willing to go into negative profitability.

So we see a good support for silicon metal prices worldwide.

So that would be my first comment.

And second is that, well, of course, we will never be running a facility at losses.

So if we see prices going down, we have proven in the past and we will, in the future, be very active in idling capacity if we go to negative territory.

What we are seriously considering, and we are looking at that, is switching additional capacity from one product to another, depending on the relatively -- relative profitability of the different products, and that is one advantage we've got as a company that none of our competitors has got.

.

Okay.

Any imminent plans on the furnace change-overs?

What -- or maybe how much capacity could possibly change over from one alloy to another?

Yes.

We're now looking at -- in the -- I would say starting September, of switching one additional furnace in Europe from silicon metal to ferrosilicon.

It's one net we would be -- actually, switching back a silicon metal furnace from ferrosilicon to silicon metal but then switching 2 furnaces from ferrosilicon to silicon metal just for better logistics.

So net-net, it would be one additional furnace on ferrosilicon versus silicon metal in Europe.

In North America, not yet, but we are, again, closely monitoring demand evolution and price evolution, and we would be considering switching one of our facilities, mainly -- maybe Selma, the 2 furnaces, from silicon metal to ferrosilicon.

That is something we have in the drawing board.

We're analyzing, and depending on the evolution, we could be doing.

Our next question comes from Vincent Anderson with Stifel.

Welcome aboard, Phil.

My first question's on capital allocation over the next 9 to 12 months.

The working capital releases and cash generation is encouraging.

But when you look at your opportunity to refinance your debt in 1Q '19, how do you plan to balance that with the cash needs of the share repurchase program, whatever additional dividends you expect you may be able to declare?

Can you just kind of what -- lay out how you expect to approach those 3 buckets over the next, call it, 9 to 12 months?

I'll allow Phil to talk about the refinancing.

What I would say is that when we look at second half of this year, we see very strong free cash flow generation.

And we are confident that with that free cash flow generation, we can undertake all that we are announcing, so about the share repurchase, the dividends and still, maybe even with all of that, be able to slightly reduce our net debt in 2018 [as a whole].

So our view today is aiming to have positive free cash flow in 2018 as a whole, and that allowing us to undertake, again, the different commitments of returning value to shareholders.

And with regard to where we are with the refinancing and what the impact on cash flows, Phil?

For the refinancing, we're actively starting those conversations now.

I think given the changed position of the company, we have lots of options available to us, and we'll continue to come back as we proceed with those discussions.

We really do expect, though, that any new debt we undertake is going to reduce down our interest expense.

If I look at our cash generation over H2, I think we have committed here to about $95 million of specific initiatives, and we're already making progress on these initiatives.

So beyond what we do every day, I think there's more being done in H2.

And then I think also if we look forward to next year, then a lot of what's happened in -- during the first half of this year, it is not going to continue on a normal basis.

I'd expect normalized CapEx to come down.

That should be in the range of $70 million to $80 million per year.

You will see those impacts we're talking about on interest costs flowing through, and you'll see -- you won't see the build of working capital as we move forward.

So very confident when you add all these things together, we're going to deliver positive cash flow in the second half of the year.

We will absolutely be targeting positive free cash flow for the full year, and we have committed to those targets we've set on deleveraging for the company.

And we try to balance that out with returning value to shareholders and looking at those other investment opportunities.

Great.

So turning over to manganese alloys.

Steel markets have been strong, and traditionally, there's been decent raw material pass-through from the ore side historically, as you've mentioned.

So just what is driving this disconnect that we continue to see?

Is there a new capacity that we haven't noticed?

Is there a volume-focused integrated producer, such as what we have in Eastern Europe just being kind of a bad actor?

What do you attribute this disconnect to?

It's difficult to tell.

I would say that there is a strong evidence of correlation historically, so we are confident that this will return to normal.

I would say that alloy prices for a long time have been discounting ore prices going down, and that is why I think they have not been reacting more than what we would have expected.

Again, there is some signals in China of silicomanganese futures going up.

I think -- again, we have -- we are very confident that we will see either alloy prices going up and most surely, in the medium term, ore prices going down.

The ore market is a well-supplied market, and we are confident that it will return to more normal levels.

And if I could sneak in just one more.

I was hoping to get an update on the KTM initiative.

Have any efficiency factors been identified?

And can their impact be quantified as to what kind of time line we could see an impact?

What plants it would be rolling out to or at least maybe not plants specifically, but product segments?

Anything along those lines?

Yes, we are monitoring, I think, the main improvements -- and we will size that.

We are -- the analytics of sizing, that is not always [specific] because it implies a number of variables.

But we are certainly seeing an improvement in energy efficiency across the group, and I'm talking mostly in silicon metal.

So we certainly have seen a very good impact of sharing best practices across the group in terms of just normal technical performance of the furnaces, and we are seeing that happening.

So there is, more or less, average, I think, of 400-kilowatt hours per ton improvement in furnace performance overall in the group.

So that is significant, and we have to run the analytics to make sure what is the economic impact.

I think the other area where we have done a lot of work and is being very interesting is electrodes.

In -- as I was saying, in the context of increased electrode cost across the world, we have been very flexible in changing from one technology of electrodes into others.

And we have benefited from the expertise we have, again, in different technologies, and that has proven to be very effective in furnace performance and in economic performance.

(Operator Instructions) Our next question comes from Sarkis Sherbetchyan with B. Riley FBR.

Phil, you mentioned the recurring CapEx for the back half of this year below the 1 half level, and I think you gave a run rate number between $70 million and $80 million per year.

Can you maybe line up the expectations of CapEx for the back half of this year and how you get to, call it, $70 million and $80 million per year on a forward basis?

Just to clarify, Sarkis, that $70 million to $80 million was on a normalized basis.

I think if we look to the second half of this year, clearly, we have the solar project, which Pedro took you through, to complete.

And we have some further spend there, that committed -- committed CapEx there is [$13 million or so].

And I think we have generally said that this year's total of the solar project in the year is somewhere around $60 million.

So it wouldn't be surprising that our total CapEx in the year is $50 million to $60 million above a normalized level.

Understood.

That's helpful.

And then I think either in the prepared remarks or the slide itself, the noncore asset divestitures' figure of about $20 million, can you maybe remind us what those pieces relate to?

Would it be the hydro facility that was on sale earlier?

Or would this be some other noncore assets?

We are looking at 3 tangible alternatives, one is almost closed and which is spending collection, I would say, and that has to do with timber farms in South Africa.

A part of that has already been realized in the first half of the year.

The remaining is somewhere around $10 million, I believe, in the second half of the year.

The rest is -- one is hydro assets that are not subject to regulatory approval.

The effective size of that is around 20 million -- sorry, 20 megawatts and, I say, effective because the rest of it is hydro assets with concessions that expire next year, so basically no value.

And I think we could get in the range of EUR 20 million to EUR 25 million for those and, hopefully, be closing that in the second half of this year.

The third alternative we are exploring is maybe a minority stake in our mining operations in South Africa, and that is something we're pursuing.

We still don't have a size.

So when we talk about $20 million, I think we may be being conservative, but we want to be conservative when we talk about cash management and balance sheet management.

And we think we could even be above the $20 million we are -- we're talking about in the presentation.

Okay, that's helpful.

And then one last one from me.

I think you mentioned some evaluation of plants, which arose from the earlier questions.

And I also believe some of the extra maintenance or plant overhaul costs we saw perhaps may not be recurring.

So I guess in that context, if you were to do a switchover in a go-forward quarter period, would there be some incremental maybe a couple of million dollars of extra expenses associated with that?

Or at this point, you don't really expect some of those extra expenses.

Can you maybe help us understand that?

That's a very good question.

I would say that the plants that are now more specific and I talked about in Europe, those are, I would say, within the normal CapEx and cost trend.

So those will -- should not be showing as exceptional.

If we do anything in North America, that could have some additional costs during the year that would show as extraordinary costs.

You're right on that.

And I don't have a sizing for that because, again, that is still in a very preliminary stage of evaluation.

I'm not showing any further questions at this time.

I would now like to turn the call back over to Pedro Larrea for any closing remarks.

Well, that concludes our Q2 earnings call.

Thanks, again, everybody, for your participation.

We look forward to hearing you -- to hearing from you on the next call, and have a great day, everybody.

Thank you, everyone.

Ladies and gentlemen, thank you for participating in today's conference.

This does conclude today's program, and you may all disconnect.

Everyone, have a great day.