Livent Corp (LTHM) 2019 Q2 法說會逐字稿

Good morning, and welcome to the Second Quarter 2019 Earnings Release Conference Call for the Livent Corporation. (Operator Instructions)

I will now turn the conference over to Mr. Daniel Rosen, Manager, Investor Relations for Livent Corporation. Mr. Rosen, you may begin.

Thank you, KC. Good morning, everyone, and welcome to Livent's second quarter earnings call. Joining me today are Paul Graves, President and Chief Executive Officer; and Gilberto Antoniazzi, Chief Financial Officer. The slide presentation that accompanies our results along with our earnings release, which includes our 2019 outlook, are available on our website. The prepared remarks from today's discussion will be made available after the call. Following our prepared remarks, Paul and Gilberto will be available to address your questions. (Operator Instructions).

Before we begin, let me remind you that today's discussion will include forward-looking statements that are subject to various risks and uncertainties concerning specific factors, including, but not limited to, those factors identified in our release and in our filings with the Securities and Exchange Commission. Information presented represents our best judgment based on today's information. Actual results may vary based upon these risks and uncertainties.

Today's discussion will focus on adjusted earnings for all income statement and EPS references. Reconciliations of these terms as well as other non-GAAP financial terms to which we may refer during today's conference call are provided on our website.

With that, I'll turn the call over to Paul.

Thank you, Dan. Good morning, everyone. I'll begin by providing a brief overview of Livent's business today. Gilberto will then discuss 2019 second quarter results, our third quarter and full year guidance, followed by select financial data and our latest capital spending plans. I will then conclude by reviewing our perspectives on current market dynamics and their implications for both the near-term and longer-term trajectory of the industry.

So turning to Slide 3. Our business performed as expected during the second quarter, with revenue, adjusted EBITDA and adjusted EPS all in line with our guidance. Revenue for the quarter was $114 million or 6% higher than last year. This performance was driven by a 21% year-over-year increase in volumes led by lithium hydroxide and partially offset by lower lithium carbonate sales and a decrease in the average price realized on both products. Pricing for butyllithium and high-purity metals were higher on average on a constant currency basis. And overall, price and mix had a negative 13% year-over-year impact on revenue while foreign exchange added a headwind of 2%, most notably from the RMB and the euro.

So now that we're over halfway through 2019, we thought it was important to review the performance of our business in the year-to-date on Slide 4. Livent increased its total LCE sold by approximately 20% in the first half of the year compared to the same period in 2018. We made a number of operational improvements in Argentina that have allowed us to increase our run rate production levels of lithium material and significantly improve our ability to operate under a wider range of conditions. As we will discuss later in more detail, we expect volumes to improve even further in the seasonally stronger second half of this year, with total LCEs available up over 30% compared to the first half of 2019 and over 20% higher than the second half of 2018.

Another big driver of this increase in LCEs is Livent's expanded lithium hydroxide network, with hydroxide volumes up roughly 50% through the second quarter year-over-year. We brought a third hydroxide line in China into production during the first quarter, taking total annual capacity in China to over 13,000 tons. We're currently operating our hydroxide network at record levels of production. Just as important, we're supplying some of our larger customers from multiple locations, demonstrating our ability to produce a consistent product across manufacturing units.

In Argentina, we continued to meet critical milestones for the 40,000-ton expansion project, and we remain on track to bring the first 9,500-ton phase online by the end of 2020. We've built onsite worker housing facilities, we successfully executed labor agreements, and we obtained critical permits from the Catamarcan government. The construction of the modular carbonate production units is on track, and we expect the first units to be shipped to Argentina later this year.

Most importantly, we recognize that we have a commitment to the local communities in which we operate, and we take this responsibility very seriously. We will continue to take a partnership-based approach to our expansion in Argentina.

Separate from our capacity expansion undertakings, we have recently concluded debottlenecking activities on our existing production assets in Argentina. We expect that this will generate as much as 1,000 incremental tons of LCEs on an annual basis. We're proud of the improvements we continue to make in a region where others have struggled to meet both targeted production capacity and quality levels, and our success has been a testament to the strength and operational expertise of our team.

There've been a few factors that have come together in 2019 to drive Livent's costs higher as compared to last year. Some factors have been out of our control, such as weather-related disruptions and foreign exchange rate fluctuations. However, demand from our customers continues to outpace our ability to produce sufficient product, in particular, the lithium carbonate needed to feed our hydroxide units. This has required us to purchase third party carbonate, and we expect to continue to do so until our first phase of carbon expansion comes online.

And with that, I will turn it over to Gilberto.

Thank you, Paul, and good morning, everyone. Turning to Slide 5. Adjusted EBITDA for the quarter of $28 million was in line with our guidance, and the primary drivers of the year-over-year decline were consistent with our expectations.

Volume growth continues to be driven by increased deliveries of lithium hydroxide to our customers but partially offset by lower carbonate sales. While a favorable price/mix was largely driven by customer mix, we also realized lower pricing on certain hydroxide and carbonate volumes sold during the quarter, particularly in China. The higher costs incurred in the quarter were largely driven by higher purchase of third-party carbonate, as we highlighted in our previous earnings calls. Foreign exchange also had a negative effect on earnings, with the euro and RMB primarily impacting revenues and the Argentine peso impacting cost.

Turning now to Slide 6. For the third quarter, we expect revenue to be in the range of $105 million to $115 million, essentially flat at the midpoint compared with last year as the trend of higher volumes and lower average prices in the first half of the year continues. Adjusted EBITDA is expected to be in the $26 million to $30 million range, and adjusted EPS is expected to be in the range of $0.11 to $0.14.

For the full year, we are reaffirming our guidance. 2019 revenue is expected to be in the range of $435 million to $475 million, slightly above 2018 at the midpoint. We projected adjusted EBITDA between $125 million and $145 million and adjusted EPS in the range of $0.56 to $0.66.

Turning now to Slide 7 and a comparison between the first and second half of 2019. Put simply, Livent's improved second half performance will come from a meaningful increase in volumes sold. And to be clear, we are not forecasting price improvements in the second half of the year. We expect to have over 30% more LCEs available for sale as compared to the first half of this year. These volume increases will be across our lithium products, but most notably in lithium hydroxide and carbonate.

More importantly, we expect to have about 3,000 tons of additional material from our own operations. This increase is driven by 3 factors: First, we do not expect the January rain event in Argentina to repeat in the second half of the year. Second, historically, we have seen seasonally higher volumes of productions in the second half of the year. These 2 factors account for the majority of the increased volumes expected in the second half.

We will also benefit from the debottlenecking actions that were recently completed and from inventory we built in the second quarter. Given our low-cost position in Argentina, this incremental volume will give us the opportunity to realize higher margins than we achieved on average in the first half of the year. We are assuming a lower average price in the second half of this year, which we expect to be partially offset by a more favorable product and customer mix. From a cost perspective, we do not expect a significant difference between the first half and the second half.

We recognize that our guidance implies a stronger fourth quarter than we have seen in the first 3 quarters of 2019. The last quarter of the year has historically been the highest-volume quarter for Livent, and we expect that pattern to repeat this year. We are already seeing rising monthly demand in certain key markets, which Paul will discuss in further detail shortly. This rise demand is also supported by specific conversations we have with our customers. Livent material is being qualified to new platforms and current indications are that customers will start taking delivery of product for these applications before the end of the year.

On Slide 8, I wanted to provide a few comments on select items of the income statement as well as give an update on our capital spending plans. For taxes, we have increased our guidance range on adjusted tax rate for the full year by 100 basis points to 19% to 23%, reflecting our latest view on the impact of earnings mix across the various geographies. Livent generated adjusted cash from operations of $67 million through the first half of 2019. We spent $74 million on capital spending in the same period and ended the quarter with debt net of cash of $39 million.

We are increasing our guidance for adjusted cash from operations for the full year by $50 million to a range of $90 million to $120 million. Our full year guidance for capital spending has been lowered by $25 million to a range of $210 million to $245 million (sic) [$240 million], which reflects an updated view on timing of cash outlays and should not be understood as a change in our expansion plans.

In Argentina, we continue to make significant progress in meeting critical milestones for our carbonate expansion project with cash outlays of nearly $50 million through the end of June. Other capital spending of $26 million in the first half of the year includes maintenance CapEx and initial investment in our 5,000 ton hydroxide expansion in Bessemer City. We expect capital spending to accelerate in the second half of this year as projects continue to progress and construction milestones are met.

On our last earnings call, we emphasized our commitment to the 4 phases of lithium carbonate expansion in Argentina, tripling our capacity to 60,000 tons by 2025. Our expansion remains a highly attractive investment that compares favorably with every other lithium project globally, given both our industry-leading experience and our cost position. We remain committed to investing in the lithium hydroxide capacity necessary to meet the demand of our customers, as shown by our Bessemer City expansion plans. We constantly evaluate and reassess our specific lithium hydroxide expansion plans based on the needs of our customers. As part of this process, we are looking to increase our operational flexibility by creating a broader global network, which involves an assessment of the most suitable geographical locations for additional hydroxide units. However, we will not -- we will only make final investment decisions when we have sufficient commitments from our customers. We have already started conversations with some of our largest customers with respect to future hydroxide capacity investments.

With that, I'll turn the call back over to Paul.

Thank you, Gilberto. I want to conclude on Slide 9 by giving a high-level update as to what the key lithium markets are looking like today from Livent's perspective. There were a lot of positive indicators in the market, with demand fundamentals better than we had expected 6 months ago. There is, though, a supply overhang that is likely to prevent prices from increasing before 2020. However, demand is growing faster than we had previously forecast and supply additions are slowing, giving us confidence that today's pricing environment does not reflect the longer-term outlook.

The market for electric vehicles continues to be very strong. Passenger EV sales globally were up roughly 45% through the first half of the year compared to the same period last year, and sales in China were up over 60% in this period. More importantly, when we combined EV sales with average battery size, we can see that the total reported megawatt hours of battery capacity sold in passenger EVs was roughly 150% higher in China for the same period, and ultimately, it is this metric that drives lithium consumption.

For the most recent quarter, reported sales of higher nickel chemistries, such as NMC 622, NMC 811 and NCA, were up 40% in China compared to Q1 2019. Not surprisingly, given this trend, hydroxide sales are significantly ahead of last year, with global sales approaching 50,000 metric product tons through the first 6 months, roughly double the same period in 2018. Independent data suggests that sales of hydroxide was over 10,000 metric product tons in June alone.

Based on this data, we are increasing our 2019 hydroxide market demand forecast to 100,000 to 110,000 metric product tons, an implied 55% to 71% year-over-year increase compared to the last year. We will revisit our 2020 demand forecast numbers once we have Q3 market data and have had more discussions with our own customers about their needs for 2020 and beyond.

We believe that one of the key drivers of this faster growth is the increased involvement and direction from global auto OEMs, which is providing greater clarity throughout the entire electric vehicle supply chain. As OEMs prepare for new vehicle launches in 2020 and beyond, they're taking a number of important steps to prepare. Many are now announcing firmer commitments with multiple battery suppliers, typically on a vehicle platform basis and have been more precise with specifications as to battery type, cathode technology and battery pack size. Just as important, they are more directly engaging in the qualification processes, not just of the batteries themselves but also of the battery raw materials. The greater involvement of OEMs has, in turn, helped the battery producers more clearly plan their own production road maps. And as a result, we're seeing more qualification activity taking place by battery and cathode producers as they look to establish hydroxide purchasing agreements for 2020 onwards.

As the OEMs get more directly engaged, we are seeing this translate into more stringent specifications for lithium and slower qualification processes as a result. A greater variety of specifications are also being requested, whether it relates to impurity levels, particle size, moisture levels or even whether the hydroxide is produced directly from spodumene or brine-based carbonate. This is resulting in a rapid proliferation of the forms of hydroxide that are needed, making it more challenging to meet customer requirements. Equally important, because of the cost and time required to qualify producers, battery companies are looking to limit the number of lithium suppliers they qualify on each platform or vehicle.

I'd like to comment on the inventory and supply situation in the market, which is having a negative impact on the pricing environment for 2019 but which also points to the likelihood that when pricing for hydroxide does recover, it is likely to do so quickly. There is continued oversupply and excess inventory of spodumene in Australia and China as Australian miners have brought on more capacity in the first half of this year. Much of this inventory is in China, but there are also meaningful inventory levels in Australia. Consumers of lithium products are expecting lower prices until this inventory has been consumed.

Hydroxide inventory of Chinese producers has been built ahead of demand, reflecting the economic incentive to run plants continuously to keep costs per ton down. However, hydroxide has a shorter shelf life and, therefore, is being pushed into the market creating further pressure on prices.

Carbonate inventory is elevated at some customers, particularly outside of China. This is due to a number of factors, most notably the temporary suspension of production of ESS batteries in Korea and the more rapid than expected move away from carbonate-based chemistries towards hydroxide-based chemistries for the EV market. However, we believe the destocking of hydroxide inventory at cathode producers is largely complete, with customers now operating with much lower inventory levels. This is not a surprise given broadly held views as to pricing trends in the coming months, but it is an important factor, in our view, that when prices do start to increase, they will do so very quickly.

The last point is connected to my final observation about the supply situation in the hydroxide market. At current prices in China, third-party reports indicate that many spodumene converters are operating at the marginal cost of production, even allowing for lower spodumene input costs, which clearly isn't economically sustainable for them in the long run. Perhaps more important, supply additions outside China are increasingly being delayed, reduced in scale or put on hold altogether. Spodumene expansions in Australia are being delayed as we believe that current spodumene pricing is insufficient to cover both operating costs and reinvestment costs for many producers. And crucially, previously announced projects to build at-mine hydroxide plants are also being reassessed as both capital cost of construction and the forecasted operating costs are simply too high for them to be viable with prices where they are today.

Not surprisingly, given the environment we are in, traditional financing for new projects is largely unavailable, while nontraditional financing sources are prohibitively expensive. While this will not impact 2019 or 2020 supply, much of the announced additions due to commence for 2021 are unlikely to start up until later, if at all.

So to conclude, while we remain cautious on the 2019 pricing environment, we are increasingly positive on the long-run pricing environment. We remain committed to adding the capacity and the capabilities needed to meet the requirements of our customers, not just today, but over the next decade and beyond.

I will now turn the call back to Dan for questions.

Thank you, Paul. KC, you may now begin the Q&A session.

(Operator Instructions) And your first question comes from Chris Kapsch with Loop Capital Markets.

Just in looking at your expectations for the second half, you mentioned that costs would be down partly on reduced VAT expenses. I'm just curious if the demand, therefore, is implied that it's picking up in China. Can you just characterize the nature of the customer there? Are those Chinese customers that previously sort of stalled out on their purchases and pushed back on pricing? Or they tend to be more of the sort of "Western" integrated battery guys that have operations in China?

Chris, it's Paul. Thanks. Look, it's plenty of both of those. I think it's important -- when we stand back and we look at the data and where we sit today, I think if you recall at the start of this year, we said the migration to high-nickel cathodes was probably delayed 6 to 9 months, and we sit here 6 months later maybe talking a lot more confidently about the demand patterns. When we look at monthly data -- and let's just be clear, the data while directionally accurate is not necessarily accurate to the ton -- we see purchases of lithium hydroxide, in China particularly, at double the levels in June than they were in January or February. And what we're seeing is a ramp-up really across the board, both Chinese and the Chinese arms of Korean and Japanese producers, all increasing their purchases as we went through Q -- late in Q2.

Now a month or 2 is not long enough for us to make a full year prediction clearly, so we're looking to see where July and August data splits out. But I think the purchases are really coming from both Chinese domestic and international battery supply chains in the second half of the year.

Okay. And then if I could, the -- your comments on sort of the qualification cycle getting more complex for some of the advanced cathode chemistries and some OEs getting involved and even specifying where the, ultimately, the feedstocks or the feeds for those cathode producers are coming from, is that something that you feel benefits brine versus hard rock? Or more specifically, is it beneficial to Livent given the steady-state production you have out of your asset in Argentina? If you could comment on that.

Yes. No, I think -- I don't expect that the entire market is suddenly going to start rejecting spodumene-based hydroxide. But there will be and, in fact, are companies out there who are looking at the broader environmental footprint of the battery and saying that from their perspective, the lower carbon intensity of a brine-based production process is preferable to them in their own product. They've essentially centered their product on that basis, around sustainability, et cetera, and it will be a small part of the market.

We also have customers that are very much requiring completely different products depending on the application it goes into. We will have the same customer, unlike a few years ago, instead of sliding us a single specification sheet across the table for their required purchases, they'll give us 3 or 4, and it will depend upon the application. It will depend essentially upon the very specific cathode that is going into the end application. And this is really driven maybe for the -- not entirely for the first time, but in a larger way by the consumer rather than by the producers. We're seeing the consumers dictate what is being required. And by consumers, I mean either the OEM or the maker of a consumable material, whether it's a power tool or a e-mobility scooter application, looking very much directly at what they expect to be in their batteries.

And so to answer your question, I think it will just create more fragmentation in the short term and more complexity. I think already at this point, we are having to stand back and make decisions about which segments of that market are we going to tailor our production processes to meet. We do have questions, given we currently operate 4 different independent lines and will add a fifth one late next year, should we run each of those 5 lines producing different specifications? These are questions we're still wrestling with, to be perfectly honest. And I think they just add complexity for all of us in the qualification process.

Your next question comes from Christopher Parkinson with Crédit Suisse.

Great. Can you just offer some further clarity and detail regarding your conviction on your ability to bring on the additional 40,000 tons in Argentina on time and on budget? Just any color on key milestones you've met, any potential bottlenecks, et cetera, just will be appreciated.

Sure. The process of bringing additional capacity in Argentina is a complex one. As you know, it's complex from an engineering perspective. It's complex from, frankly, a logistical perspective. And there are sensitivities to operating in Argentina that are unique to Argentina. It's a remote part of the world, and we have communities and political will in that country that's maybe different to other places.

I think the key challenges to us have largely been around, and for most people, I think, have been largely been around the logistics of infrastructure in that part of the world. And once you get into that process of infrastructure building, you have to have a very close partnership with the local authorities because you need permits, you need permissions, there are local laws that have requirements as to how much local labor and local content is going into there, there are requirements that you have to put in place with regard to maintenance of ancillary infrastructure, about training of local employees. These are all real commitments that we make and that we have to make. And our focus has largely been ensuring that we have the correct relationship with all the local parties so there are no misunderstandings and that there are -- there is a real sense that this is a project that is being developed for the benefit of all parties involved.

We have received in the first half of the year the environmental permits, the construction permits that we need to stay on track for an end of 2020 go live on the first carbonate unit. We've also separately been able to put in place some of the key contracts that we have in place, that we need to put in place. These are either construction contracts with contractors in Catamarca or they are construction contracts for fabrication yards in China to build the actual carbonate units themselves.

So when we go down our list of key milestones, they fall into multiple areas, and we are on track. Every one that really matters to get us to an end of 2020 go live, we're on track for, we've hit so far this year.

And obviously, there have been a lot of moving parts during the first half of this year, but just as we're obviously a longer-term story, as we're heading into 2020, can you just kind of break down your cost structure just on a general per ton basis, just addressing the differences that derive from the Chinese VAT, third-party product versus internally procured tons and the North America ramp? Just any color on just how we should think about kind of the normalized environment for '20 and '21.

Sure. Look, I think in the normalized environment, first of all, we would expect -- I think we maybe talked about this more than most -- that we expect local production to serve local markets generally. What does that mean? We wouldn't expect to be producing in China primarily for export. So we would expect that VAT leakage to largely go away over time. As we see more and more migration of demand into China, it's highly likely that as we get through 2020 and into 2021 that the volumes that we are exporting now to China will start to decrease and, therefore, the VAT will go down.

Our general level, I think we've talked about this in the past a few times that for carbonate in Argentina, it is a low-cost production facility. We cannot debate depending on where prices and royalties in Chile are, et cetera, who is quite the lowest. But I think the brine producers in South America are all largely producing lithium carbonates somewhere in that $3.50 to $4 a kilo range. And it really doesn't change fundamentally, and that will be the case going forward and will continue to be the case for our expansions as well.

Your next question comes from Kevin McCarthy with Vertical Research.

Yes. I wanted to ask about your pricing expectations. I think Gilberto commented that you do not need price in order to achieve your 2019 guidance. What are you assuming for price? In other words, do you need price stability from here in order to achieve that level? Or what are your expectations for the next couple of quarters?

Sure. Look, I think you have to break our business down into pieces that are exposed to the forces that you guys all see out there and the pieces that are not. We still have a significant business in the butyllithium, high-purity metal and other specialties area. Pricing in that market is likely to be higher, and this is all largely contracted volumes with customers. These are typically customers that have been with us for many years, if not decades, and have our product largely specified into their own production processes. And we're seeing pricing trends in the second half of the year in that business to be broadly up, slightly up low single-digit percent, but certainly favorable in the second half versus the first half.

We have, as you know, a relatively low exposure to carbonate. I think our expectation is that carbonate pricing, while we will sell more in the second half, is going to be down first half to second half. I think a lot of the fall in pricing in carbonate in the markets we're selling into that is outside pre-agreed, long-term contracts, in market-based pricing has largely happened. But we do expect some price pressure still to be applied into the carbonate market, and that is factored into our guidance.

Hydroxide is a different story. We have 2 areas of demand. We have the contracts that have been in place for a few years. Those contracts, by the way, continue to perform. Customers continue to take the volumes. The pricing remains in line with those contracts. But we also have a higher proportion than we are comfortable with going into markets that are more subject to market-based pricing. Our expectation is that the average realized price that we see in those contracts for hydroxide and in our forecast, it will be down in the second half compared to the first half. It will be down at least double-digit percent in our guidance expectations.

We do have a mix benefit because some of our contracted sales in the first half have been at contracted prices that are at or below market prices today. Just the nature of those long-term contracts, as we've mentioned in the past, is they're already relatively low-priced contracts. So actually, we have a bit of a customer mix shift benefit going on in our favor. But when we think about sales of hydroxide in the second half, we're certainly forecasting a decline in pricing in the guidance that we've given.

That's very helpful, Paul. I appreciate the color there. And secondly, I wanted to ask you about China, I guess along a couple of fronts. How do you think escalation of the trade dispute will affect supply chain, if at all? And then what is your RMB sensitivity?

So I think we mentioned before, we have no U.S. to China trade in either direction. All of our China product doesn't touch the U.S., and none of our U.S.-produced product goes into China. So on a direct tariff basis, it doesn't have an impact on us. But clearly, there are significant concerns that we have as to what it does in other areas. You mentioned the RMB. We do have exposure to RMB. About 25% of our revenue is into China, but we have very little cost in China, so a depreciating RMB is a headwind for us in terms of profitability.

I think from a supply chain perspective, I think we're seeing it already. I think you heard us mention that we expect our production process to become more local, that we will produce, because of the way we convert carbonate, we will convert carbonate to hydroxide. We will increasingly produce locally for local markets. I think we will likely see and already are seeing a similar trend. I think anybody who is expecting to produce cathode materials in China or battery materials in China and export them to battery assembly or production facilities outside China is likely to revisit that. I think it likely, in fact, move more of the battery production supply chain into China and do some more quickly.

Your next question comes from the line of P.J. Juvekar from Citi.

Paul, on the last quarter, you talked about high-nickel batteries like 811 taking longer to get commercialized. And you also mentioned that some cathode guys are going to older chemistries, like LFP. Can you give us an update on that? Because this quarter, you sounded a lot more upbeat. So all in all, what changed in the quarter?

Well, first of all, as I mentioned, basically, time has passed. And the reality is that the shift to 811 remains a when, not if, conversation with every customer. The shift to hydroxide as the basis for EV battery chemistry remains a when, not if, conversation. And so timing is a variable that I think a lot of people seem to miss when they take the statement and that situation and assume it will last forever, when what we really have is a very, very dynamic environment.

We believe -- and again data is not that easy to get hold of, and so we do our best with what's out there and with some of our own conversations with customers. But it feels to us that high-nickel chemistry is in the first quarter of 2019 and in the second quarter of 2019. And by that, I mean purely 811 and NCA. In China production perspective, it was about 10% of total production.

Now that's important to understand because total production of those cathodes was up 25% Q1 to Q2. So we're seeing this shift. From relatively low percentages, we're seeing a shift in production of high-nickel cathode materials. We're seeing those cathode materials, first and foremost, finding their way into non-EV applications, where maybe the safety environment feels a little different. We are seeing what I believe to be the dying kind of growth story in LFP or NMC 111 or maybe even 532 chemistries. While they will continue to have large applications, whether it's in stationary storage or in certain nonpassenger EV applications, it feels like the technology road map for passenger EVs is moving more quickly and more rapidly towards high nickel chemistries.

We're seeing 622, for sure, grow more quickly than anything, up about 45% Q1 to Q2 based on the data we have. That's a meaningful shift. Some of that product is hydroxide-based. Most of it is not and it's carbonate-based. We're also seeing more and more blends. You have to bear in mind much of these cathode materials are blended together, so we're seeing more 811 content in batteries. And so I think what we're basically seeing when we look at the data, a combination of the pull, if you will, of the specifications being given to the battery producers by the OEMs as well as a continued ramp-up in the learning curve, not only of making the 811 materials, but also putting them safely into battery applications, continues to drive higher growth in those product areas.

That's helpful. And my second question is on the spread between carbonate and hydroxide. I mean that's important to you since you buy some -- a little bit of it in carbonate, but sell mostly hydroxide. That spread compressed, I think, in second quarter. How do you see that playing out in second half?

So I think you may have heard me say a few times that I expect 2 disconnects to happen in the future that maybe are starting to happen already. One of them is the disconnect of China from the rest of the markets around the world; and the second one's a disconnect in pricing between carbonate and hydroxide. They won't become completely disconnected. But generally speaking, they both are driven by their own independent supply-and-demand characteristics, and so you have to look at them a little separately. I think when we had a market where, fundamentally, most hydroxide was converted from carbonate, you could understand the rationale for a premium of hydroxide over carbonate. What you have today is a situation where we have more and more producers going from spodumene to hydroxide. And at today's prices, it's probably cost parity to maybe even slightly lower cost to produce hydroxide than it is to produce carbonate.

I think what we have in the market today, depending on where you look, is, generally speaking, not a premium in place at all, not much of a premium in place at all today. It's really though driven by supply characteristics. I think we've been, frankly, surprised by the pace at which hydroxide capacity in China has come online. I don't see that pace -- it just cannot continue at that pace relative to demand. But at the moment, what we have is a product in lithium hydroxide, which is being produced to certain specifications, and a Chinese producer with all the pressures they face to run their own balance sheets and financing with an incentive to sell that product even if it's a relatively low price, low margin.

And so I think that disconnect is starting to happen. I think it will continue to happen. Where it goes is a completely different and much more difficult question to answer for you. But I do think we will find a greater independence of pricing characteristics between those 2 markets in the future.

Your next question comes from the line of Bob Koort with Goldman Sachs.

This is Dylan Campbell on for Bob. When I look at your cash flow from operations and compare that to EBITDA, it looks like CFO to EBITDA was about 120% in the first half of the year, and kind of the implied guidance for second half of the year is maybe just shy of or slightly below 50%. What kind of drives that decline in kind of cash flow conversion from EBITDA?

Look, I think it's pretty straightforward. I think it's a working capital question. We inherited as an independent company what I'll loosely call some complex supply arrangements with our Chinese partners. And we spent the first part of the year simplifying those and taking working capital out of that part of the supply chain. I think in the second half of the year, with our expectation of higher sales, we're going to see that reverse again and a more normal build of working, a timing-driven build of working capital largely in the form of receivables in Q4. So it's really nothing more than the shift of working capital patents that we've talked about.

Got it. That's helpful. And in the past, you've given pretty interesting color in terms of marginal cost for, I guess, a nonintegrated lithium producer. I think the last price something just above $10,000 a ton for lithium carbonate. But I'm curious, how does that compare with your view on kind of the marginal nonintegrated producer?

Yes. The math is pretty straightforward. The single biggest driver of cost is spodumene pricing, right? And so when you get spodumene pricing down by $100 a ton, that equates to almost $1 off their marginal cost of production. There are other variables which matter, including the quality of that feedstock and the consistency of it, which impacts our operating rates, which again factors into that pricing. But generally speaking, today, we probably see the marginal producer of both hydroxide and carbonate in China, and this is verified by independent observation, somewhere between $8 and $9 a kilo or $8,000 and $9,000 a ton is the marginal producer of a nonintegrated producer.

Your next question comes from the line of Steve Byrne with Bank of America.

Paul, I wanted to talk a little bit more about this shift in cathode technology to 811. Would you say that the price premium for hydroxide over carbonate is at all relevant in the pace of that shift? And would a lower premium accelerate that shift, in your view?

I don't think it's relevant at all, to be perfectly honest. Lithium still remains, while an important, by no means the most important part of that cost structure for 811, I think availability of qualified lithium hydroxide is going to continue to be a challenging question for some of the producers. I think sometimes it's easy to get caught up in the moment. And while I wouldn't take every statement given to me by a customer at face value, we have customers turning around to us and saying, "I needed somewhere between 4 and 8x as much lithium hydroxide in 2020 and I ordering 3x as much again in 2021." And that creates a different conversation, but price is important. Price is critically important. Bear in mind that a cathode producer, generally speaking, is passing on the cost of his battery raw materials to the ultimate consumer. He's passing them on and through, in this case, to the OEM. So the OEM is driving that conversation as much as anybody else. And the OEMs' conversation has tended to be about market-based pricing. We want to make sure we're not paying a price that is different to the prevailing market price. It's an interesting dynamic because different OEMs come at this from different angles, and some see that as a real challenge to their own profitability because they don't understand where is the incentive in the battery chain to take the costs down. All costs land on the OEM, and so don't like index pricing. Others, frankly, just prefer that transparency. What we're not seeing is any conversations that say, "Because of the price of lithium hydroxide and our expectations, I mean, we're going to delay the introduction of the cathode chemistries." I've never heard that conversation raised.

Okay. You talk about your volume outlook for the second half being kind of back-end loaded and could be significant volumes taken by year-end. Can you comment on how much of your -- of the volumes implied in your guidance are actually contracted for, for the second half? And what's the risk that some of that volume could slip into 2020?

Yes. So some of that volume could slip into 2020. Again, I think, as I mentioned, what we're seeing is an acceleration in demand indications from customers, not a delaying of them. But we have more conversations with customers where we've been talking about 2020 who are now starting to say, we may bring that forward 1 month or 2 or 3, and some of that demand will, in fact, fall into Q4 of this year. How much and how certain that is, is still to be determined.

I think from a contracted perspective, for the second half of the year, when we look at our total revenue expectation, it's largely consistent with the first half, where between 20% and 30% of our volume of our revenue in the second half of the year is exposed to either demand or pricing variability depending on market conditions.

Your next question comes from the line of Alex Falcao with HSBC.

I have a question regarding your yearly guidance. If you think about your third quarter guidance for EBITDA, that implies a dramatic increase for fourth quarter. I just wanted to know if that's coming solely on the back of the increasing production for in-house production or it also comes from pricing.

So I think as I mentioned earlier, our forecast for the second half is a meaningful decline in price assumptions in that forecast. So maybe let me help just a couple of key metrics for individuals. Most of the additional volume we have coming online will be available to us in the fourth quarter, so we recognize that we are asking a lot of ourselves and our view with regard to those fourth quarter assumptions. But just to put a few numbers around it, we'll have about 3,000 tons of additional product from our own production. And I mentioned earlier what our cost of production is in Argentina. At today's -- or at our forecast pricing, we'd expect to make somewhere between $5 and $6 a kilo of additional contribution on product produced in Argentina sold into today's market environment. On top of that, we have somewhere in the region of $7 million -- $6 million, $7 million, $8 million of tailwind beneficial from mix, customer mix. Product mix, too, is an important shift in the back end of the year as well as we talked about we expect to be -- have less value or cost leakage from VAT. And so what we're seeing is largely a volume-driven step-up in performance in Q4, helped out around the margins by favorable mix and slightly lower cost in certain areas.

Okay. And following up in that, I think that could explain also the lower CapEx and lower cash flow generation as well, right?

So the CapEx is really just a timing question, right? I mean I think, as I'm sure you guys have seen in many other projects, it's easier to determine what the total capital cost of a project is than it is to predict the day on which the cash goes out the door for that capital project. We have a history in this business, as many similar organizations do, of perhaps over forecasting or forecasting cash flows to happen sooner than they actually do. As we sit here now in August, we have much more clarity over when the timing of spend will be on CapEx. It is not, though, an assumption that the total cost, capital cost will be lower than we previously said. It is still absolutely in line with previous forecasts for capital cost.

The cash flow, the slower pace of cash flow generation, as I mentioned earlier, really is with regard to the impact of inventory buildup that will happen in the back half of the year -- or sorry, working capital buildup as a result of higher sales in Q4 where, frankly, the collections for that are likely to take place in 2020 rather than 2019.

Your next question comes from the line of Joel Jackson from BMO Capital Markets.

Paul, maybe you could start by following up on some comments you made a little while ago. You talked about how the specs for hydroxide are changing even whether customers want it produced from brine or from spodumene. Could you elaborate with some more specifics on that, please?

Sure. Look, I think there are certainly -- as I'm sure you know, there are multiple technologies out there with regard to cathode chemistries. And so we'll talk about 811, but not everybody's 811 is the same. We have variations on NMC, variations of NCA. We have, frankly, a merging of NCA and NMC technologies. Particularly once you get into 811, you start to see more similarities than differences perhaps between NCA and NMC, and so we see a merging going on.

Well, I tell you is because, simply put, many battery producers and, importantly, cathode material producers are talking to the ultimate consumer, the OEM, and looking to differentiate their product, making sure that their product is specified into ultimately the most important platforms, the most important vehicles. Part of that conversation continues to be either supply chain security, supply chain exposure as well as what is the broader environmental footprint of an electric vehicle. And we've seen lots of press around how the battery materials themselves are the single biggest contributor to the fact that an electric vehicle when it's finished being built actually has a higher carbon footprint than an internal combustion engine.

And so there is a push. We can debate how broad, how prevalent and how accelerated that push is to look back through the supply chain at the sustainability of operations. And some customers, particularly European-based customers, are moving more quickly to the notion that they do not want to be exposed to the most carbon-intensive method of making lithium and would, therefore, be not only preferring, but in some cases, restricting their purchases to only hydroxide that can demonstrate that it has got a low carbon footprint. Today, that is essentially entirely carbonate to hydroxide-based product where the carbonate itself is produced from brine. Again, it's just one factor at work. I wouldn't for one moment say this is driving a fundamental shift in the market, but I think it does reflect the complexity of the conversations that are happening in the lithium hydroxide supply chain today.

And just following up on other questions, you kind of danced around this a little bit in giving questions -- answers to people on this. But just to understand, you talk about 20% to 30% of volumes or revenues in the second half of the year that could be exposed to demand or price variability. So just to clarify, is that 20% to 30% of revenues including butyllithium? Or is that 20% to 30% of just hydroxide volumes?

And then you talked about on the call a few months ago that the risk really was more on the volume than price, that if the customers for these -- the non-take-or-pays took the -- they would take it or they wouldn't, but they took it to be at that price contract. Could you just clarify a little bit on that, what's going on now please?

I'll try. I think one of the realities of our business in the first half of the year is that our challenge, frankly, was not demand-based. Our challenge was price-based. And with insufficient produced -- product produced, enhanced carbonate, that is, it was not economical to buy carbonate, convert it to hydroxide and sell it. And so what was happening is, yes, sure, we had customers that were willing to take product from us as contracted but not at the right price. And once we looked at that, we just, frankly, could not make money selling to them at those prices. And so I think it's fair to say that, that price exposure has been the biggest negative impact on us than anything around demand. I'll keep making the point, our problem in our business today is a supply based driving pricing down, not a lack of demand driving pricing down.

I think in the second half of the year, that factor just fundamentally changes because producing carbonate at $3 to $4 a kilo is a very different economic conversation for us internally about what we do with that relative to buying carbonate and being forced to find a home for it profitably. So it's -- customers did buy significantly less in the first quarter. Again, the data that we have, I'll keep qualifying it because there's lots of data out there that we all have to try and reconcile sometimes, but from our best reading of the data, purchases of lithium hydroxide in the first quarter were not much different than they were in Q4 '18 or Q3 '18. Purchases of hydroxide in Q2 have been significantly higher than the previous 2 quarters, significantly.

And so we've seen the customers continue to buy. We've had individual customers who were buying a couple of hundred tons of hydroxide a month put bids out at 1,000 tons of hydroxide in the last couple of months per month. So we cannot debate why that's happening. But from our perspective, the customers did not stand by their commitments to us on price. They are continuing, in our view, or accelerating their actual volume purchases.

Your next question comes from the line of Aleksey Yefremov from Nomura Instinet.

Just a question on your third quarter guidance. Will you still be restricted on volume in the third quarter? And related to that, is Argentina currently back to normal operations?

We will be restricted on volumes for most of Q3 still. We took the plant down in Argentina to do some debottlenecking in July, so clearly, we've lost some volume. Our process in Argentina, as I'm sure you know, is not a simple one, and there's a time lag in it. We can build inventory of product, but it is not saleable product. It's either in the form of brine or a lithium chloride basis, and we can then convert that into the carbonate once we're up and running. It takes us a period of time not only to do that, but then we have to get the carbonate to the right place. This is largely seaborne material, and so it can take a few weeks to get to the right place. What that means is the real ramp-up in available volumes, while we stopped producing it today and are now producing at higher rates today, it won't really be available for sale until the very end of Q3 and into Q4.

Understood. And, Paul, thank you for commentary about demand dynamics in China. So I'm trying to square the growth in demand in the second quarter versus sort of what prices are telling us. And how did the 2 reconcile? Was it an inventory destock? And a question related to that, do you see -- do have any read in what's happening in July following the change in subsidies in China? Has the demand picture changed a lot versus what you saw in second quarter?

Look, there's no doubt that the supply situation is far more bearish than the demand situation. We've seen -- our best estimate is that, as I said, about 100,000 or so, 100,000, 110,000 tons of hydroxide is going to be purchased in 2019, but we think about 140,000 tons may be produced. And so not only do you have a situation where there's a lot more volume available into the market relative to demand than we'd expected, this -- I think the single biggest driver of pricing today is a customer who, first and foremost, has destock. Why would you hold hydroxide inventories when, a, it has a finite shelf life, and b, you see the price going down in the future.

So I think we've seen destocking at the cathode material producer of hydroxide. But we've seen, based upon expectations as to spodumene prices, there's been a decline in marginal cost of production at that spodumene-based hydroxide producer. So put yourself in the minds of a rational consumer of lithium hydroxide. You're going to destock the inventory you have today. You're going to defer purchases to the future because the price will be lower in the future than it is today.

If you look at the spodumene story and see a continued increase in spodumene output, you don't see pricing of spodumene recovering anytime soon. So you don't have a huge incentive to do anything more than buy what you need in any given month. And that, frankly, is I think what we've seen going on. Meanwhile, supply comes in, in larger lumps. There's always an incentive with a, particularly on the spodumene-based unit, to just run them, to run them flat out and produce the product. Unlike carbonate, though, you're going to start asking questions how long can I hold that hydroxide as a producer of hydroxide. The longer it sits, and particularly if it's not in a temperature-controlled or humidity-controlled warehouse, the more it wants to turn itself back into lithium carbonate. As you start to head down that path of moisture and clumping in the material, it starts to fail specifications of customers. And so you have a supply that is coming on that is being rather immediately being pushed into the market. And any rational consumer of lithium hydroxide today is looking at that and saying, "I'm just going to buy what I need as I need until the situation changes."

Your last question comes from the line of Mike Harrison from Seaport Global Securities.

Just looking at the guidance, you have a $20 million range for the full year for EBITDA, but it's kind of a $4 million range for Q3, and that implies a $16 million range on Q4. Can you maybe talk about some of the variables or uncertainties you see going into Q4, maybe where visibility is relative to what you'd like at this point?

No. Look, I think it's a valid observation. The truth is that when we look out, we have 2 major variables impacting how -- whether we deliver on guidance or not. One of them is our ability to deliver those volumes that we just talked about. This is Argentinian brine-based production. This is not easy to do, and it can be upset pretty easily by various factors. So I would say, first and foremost, we have to deliver on that. I would not -- I'm not suggesting it's a challenge. But I'm equally not suggesting it's a slam dunk, and so it'll make a big difference to us. As we just mentioned, the margin on self-produced product is significantly higher than anything else.

The second is the pricing environment. We've assumed in here, as I said, pretty significantly -- significant reductions in average realized prices in the second half of the year. We've continued to be, I wouldn't say surprised, but I think there've been some price movements that have been maybe higher in lithium hydroxide than one expected. We've seen more stability, by the why, in carbonate than we expected and less stability in hydroxide. There's no doubt that we would -- that we continue to have a higher than we would like risk with regard to those 2 variables in the second half of the year.

And then my second question is regarding the production of hydroxide from some of these Chinese converters in the context of the increasing specifications. Just wondering what these converters are capable of in terms of meeting those tighter specifications? How difficult is it? And over time, are they going to improve capabilities? Or do you feel like you're going to be somewhat unique in your ability to meet these increasing specifications?

It's an interesting question. Here's my take on this one. I do believe that generally speaking, meeting specifications is something that most Chinese producers will ultimately be able to do. I think today, some of our strongest competitors are Chinese producers of lithium hydroxide who produce a very high-quality product.

I think the bigger question, frankly, is more on the consumption side. We've seen over and over again when supply is tight, specifications get relaxed a little by consumers who just need the product. When supply is a little looser, like it is today, we see those specifications tighten, and we see greater demands being placed on us by our customers. And I think that trend will continue.

I wouldn't underestimate, look, if you're producing from a single plant and you have to produce 5, 6, 7 different specifications, you have to produce to -- and by the way, it isn't simply the case of going to the lowest because some of these are physical properties that do vary depending on how you produce it. It isn't clear to me that every producer will be willing or able to produce the breadth of material that every customer needs, and so I think that's going to have a factor.

I also think, by the way, it's not easy for a battery guy to manage these supply chains. And what we're seeing increasingly is they want 2 or maybe 3 suppliers only into each platform. And so I think we'll see more and more producers of lithium hydroxide aligning with specific platforms and specific customers and producing to those specifications. So I think this is a supply chain that has a good degree of evolution still to go through.

And I will now turn the call back over to Daniel Rosen for closing remarks.

Thank you. That is all the time we have for the call today. We'll be available following the call to address any additional questions you may have. Thanks a lot, and have a great day.

And that's all the time we have for today. This concludes the Livent Corporation Second Quarter 2019 Earnings Release Conference Call. Thank you.