By Jeffrey Stafford
Tesla's (NASDAQ:TSLA) release of details on its proposed Gigafactory to make lithium-ion cells for its future electric cars sent waves through the lithium market, with stock prices for narrow-moat producers Rockwood Holdings (NYSE:ROC) and Sociedad Quimica y Minera de Chile (NYSE:SQM) reacting positively. Tesla's announcement does not change our view that lithium prices are likely to see pressure over the long run, as we think supply through the end of the decade will be more than sufficient to meet one of the strongest demand profiles for any industrial chemical. Operating from a low-cost resource base in Chile, Rockwood and SQM have the ability to invest in their economic moats if Tesla can bring its third-generation car, the Model E, to the market at a price and driving range that lead to increased adoption rates.
Tesla expects its Gigafactory to have capacity to build 500,000 pure-electric vehicles, or EVs, by 2020. Tesla builds pure-electric vehicles as opposed to other vehicles that include electric powertrains, including hybrid electric vehicles, or HEVs, and plug-in hybrid electric vehicles, or PHEVs. There's no doubt that Tesla's executing on its ambitious plans is a positive for global lithium producers as Rockwood estimates that a pure-electric vehicle uses roughly 25 kilograms of lithium - more than 7,000 times the amount needed to produce a cellphone battery. For the Tesla Model S, lithium use has been pegged at closer to 40 kg per vehicle. That said, pure-electric vehicles are growing off such a small base that the impact on lithium demand - at least to the end of the decade - is somewhat limited.
Our demand model for lithium assumes that in 2020 EV sales will account for less than 2% of global auto demand. This translates to EV lithium use accounting for only 15% of global lithium demand by 2020. If we assume that Tesla's announced vehicle capacity of 500,000 by 2020 is totally incremental to our forecast of roughly 1.5 million pure-electric units - that is, we simply add 500,000 new EV batteries to our forecast - our lithium demand compound annual growth rate from 2012 to 2020 increases only about 1%, from about 8% to 9%. We forecast industrywide EV sales of about 1.5 million units in 2020. This would certainly be a very bullish take on Tesla's battery plans, as we had implicitly built rapid Tesla growth into our prior EV forecast. As such, we're essentially leaving our long-term demand outlook for lithium unchanged, as it has already taken into account swift growth in EV production and sales.
Li-Ion Batteries Should Command Small but Meaningful Share of Auto Market
Although we are not materially changing our lithium forecast, Tesla's Gigafactory plans increase our confidence that lithium-ion batteries will command a small but meaningful place in the transportation market next to internal combustion engines and other powertrain options over the long run - even if that takes a decade or two. We see the substantial investment by Tesla - the price tag of the Gigafactory will come in around $5 billion - as a strong indication that CEO Elon Musk believes the economics can work for larger-scale all-electric vehicle production. If this is indeed the case, then we would expect other large automakers to eventually enter the market with competing products. This scenario would portend a sharp run-up in lithium demand after 2020.
A variety of factors make EV adoption rates (and thus lithium demand) hard to predict, which we think justifies our very high fair value uncertainty ratings for Rockwood and SQM. At the forefront of questions of EV viability are battery costs, which make up a substantial portion of EV vehicle production costs - in some cases up to half of production costs. Tesla is betting that it can reduce battery costs by as much as 30% by taking advantage of economies of scale at the Gigafactory. Lower battery costs would make a long-range EV affordable and practical for a much larger swath of the car-buying public.
Using the often-whispered price target for Tesla's Model E of $35,000 and the Department of Energy's Vehicle Cost Calculator, we estimate that the cumulative cost of ownership for a hypothetical Model E and a 2013 four-cylinder automatic Honda Accord would be indistinguishable at around $60,000 over 10 years, with the Tesla marginally more expensive. This analysis assumes annual driving distance of 12,000 miles split about 50/50 between city and highway, $3.50 per gallon gasoline, about $0.10 per kilowatt-hour electricity, and a 30 KWh per 100 miles rating for the Model E.
Vehicle purchase price and operating costs aren't the only issues that will determine adoption rates for the Model E and other EV offerings. For example, convenience factors including the prevalence of road charging stations, charging times, tax credits and vehicle range will all play big roles, with each of these issues potentially holding back EV adoption rates. Further, hybrid electric vehicles could still trump the Model E on cost of ownership later in the decade. For example, a 2013 Toyota Prius Hybrid bested our hypothetical Model E using the assumptions in the aforementioned scenario. Our analysis also leaves out potential fuel efficiency gains by internal combustion platforms during the time it will take the Model E to make it to market. All that said, the Model E could do considerable work in closing the gap between ownership costs that exist today between the Model S, which sports a base selling price of $70,000, and cars that appeal to a broader customer base.
Tesla's success also matters for lithium companies, because with increased range capabilities, Tesla's vehicles use more lithium per car than current EVs such as the Nissan Leaf. While we expect Tesla's future offerings to require less lithium per car than the high-performing Model S, average lithium content per battery across the industry should increase if Tesla grabs a higher and higher percentage of EV market share.
Tesla's Success Would Allow Rockwood and SQM to Invest in Their Moats
While the Tesla announcement does little to change our view of the lithium market and pricing through 2020, we believe an opportunity exists for Rockwood and SQM to expand their lithium production beyond our five-year explicit forecast period. We think these investments would be value accretive for both companies as the firms can invest in their lithium operations at attractive rates of return. Rockwood and SQM have advantages over new entrants in expanding lithium capacity. For example, Western Lithium (WLCDF), a potential new entrant, controls a greenfield project in Nevada with Phase I capacity of 13,000 metric tons per annum and a capital cost of $248 million, which works out to about $20,000/mt. To double capacity to 26,000 mtpa, which is essentially the equivalent of a brownfield project, Western Lithium estimates it will spend only $161 million for a cost of about $12,500/mt.
We also believe Tesla may look to form a strategic supply agreement with either Rockwood or SQM, or even both. While Li-ion battery costs can make up a large portion of total electric vehicle production costs, the actual cost of the lithium carbonate or lithium hydroxide in the battery is minuscule when compared with total production costs. Rockwood's CEO Seifi Ghasemi has said that the Model S uses almost 100 pounds of lithium carbonate. At a battery-grade lithium price of $6,000/mt, that equates to only about $135 of lithium for a vehicle that sells at a base price of about $70,000 (and can approach $100,000 with options). That's not to say that lithium is not important to Tesla, since a Li-ion battery requires it. Given these dynamics, it would not be all that surprising to see some sort of strategic supply agreement between Tesla and a major lithium producer. Tesla would need a ballpark figure of 12,000 mtpa to produce its 500,000 batteries at the Gigafactory, assuming 25 kg of lithium per battery. Those 12,000 metric tons represent about 8% of today's demand, and this estimate could prove to be low. Recently, when Ghasemi was asked to estimate the lithium demand from the Gigafactory, he quoted a number of 25,000 short tons or 22,680 metric tons. Ghasemi used a lithium content of roughly 45 kg per Tesla battery, which we think is optimistic, but nonetheless, we believe Tesla could make up a very material portion of global lithium demand by the end of the decade. Therefore, if it expects to demand a good portion of the total lithium market, Tesla may want to lock in a supplier and may be quite price-insensitive, given lithium's low proportion of total per-car production costs.
While it's difficult to predict what a supply agreement might look like between Tesla and a lithium producer, we can frame some numbers that will give investors an idea of the potential impact to Rockwood or SQM. The 500,000 EV batteries Tesla hopes to produce by 2020 translates to about 12,000 metric tons of lithium carbonate, using a conversion rate of 25 kg of lithium carbonate per battery. If we use Ghasemi's estimates of lithium per battery, then the impact could be much larger, at more than 20,000 metric tons. This compares with Rockwood's current lithium carbonate capacity of roughly 50,000 mtpa after expansion projects are completed in 2014, and roughly the same capacity from SQM. Both companies can expand lithium capacity relatively easily through brownfield projects. In fact, our SQM model already includes an expansion to about 60,000 mtpa that has not yet been given the full go-ahead by the board and management. If demand is available to support expansions, we would view a Tesla-related expansion by either Rockwood or SQM as a moat-widening exercise, as both companies generate nice returns on brownfield expansions. That said, our lithium volume forecasts for Rockwood and SQM already implicitly include strong Tesla growth, so the incremental expansion opportunities for the two companies is less clear.
Further, it's not a foregone conclusion that Tesla will partner with these two low-cost industry giants if it opts for some form of partnership. Tesla is reportedly considering Nevada, Arizona, New Mexico and Texas for its new battery plant, and it could back the greenfield project of Western Lithium, which controls assets in Nevada. The production numbers the company has cited could satisfy essentially all of Tesla's demand through 2020 if both project phases are completed.
Chilean Assets Bring Cost Advantages
In addition to enjoying low capital expansion costs, Rockwood and SQM sit on the low end of the production cost curve. The companies' cost advantages stem from Chilean assets in the Salar de Atacama. Salt brine is pumped from just below the surface and then dried by the sun's rays in massive evaporation ponds. Marginal-cost producers of lithium either extract the element from salt brines with lower concentrations or mine spodumene rock and process the mineral to make lithium carbonate. Both methods produce lithium at significantly higher costs than the Chilean brines used by Rockwood and SQM. Processing spodumene into lithium carbonate involves grinding, heating and dissolution, which incur higher costs than brine production. A dated, but we think still illustrative, 1990 study by P. Pavlovic estimated that spodumene-based lithium production in North Carolina was more than twice as costly as Atacama-based lithium production. We think Rockwood and SQM are at the low end of the industry cost curve and can produce lithium for less than half the cost of marginal producers, many of which produce lithium carbonate in China from Australian-mined spodumene.
Research by Paul W. Gruber et al. published in Journal of Industrial Ecology suggests that the Salar de Atacama not only is the largest developed lithium brine deposit in the world, but also holds the highest concentration of lithium. Atacama's average lithium concentration (0.14%) is much higher than that of large developed brine deposits in China (Qaidam, 0.03%; and Zabuye, 0.068%) and Argentina (Hombre Muerto, 0.052%). The higher lithium concentration helps put the Salar de Atacama at the low end of the industry cost curve.
While the majority of Rockwood's lithium production comes from Atacama, the company also produces a portion of its lithium in the United States at the Silver Peak brine deposit in Nevada. Silver Peak's lithium concentration (0.02%) is much lower than Atacama's. We believe Silver Peak production pulls Rockwood a little way up the cost curve, potentially above SQM, which produces lithium exclusively out of Atacama. That said, Rockwood's management has stressed that the company is "absolutely the lowest-cost producer of lithium and lithium compounds in the world."
Vehicle Batteries Are a Big Driver of Lithium Demand, but Growing Off Small Base
Lithium is widely used in Li-ion batteries for vehicles and personal electronics, but battery demand currently accounts for only 22% of global end use demand. We expect this percentage to increase as the growth rate for lithium personal electronics battery applications will probably be much higher (10%-15% per year) than for other uses such as ceramics and glass (2%-5%). With incomes rising in China, India and other developing economies, demand for personal electronics is set to grow. However, the big potential growth driver for lithium is the element's use in rechargeable batteries for all-electric and hybrid vehicles. Electric cars are generally divided into three buckets: all electric, hybrid electric and plug-in hybrid electric. The amount of lithium used in each type of vehicle varies significantly.
While electric cars have been on the market for some time, with the Toyota Prius the first major commercial success, lithium-ion technology has not grabbed a large share of market as of yet. The dominant technology in HEVs has been the nickel-metal hydride battery used in the Prius. Despite NiMH's leading position, Li-ion batteries stack up favorably in a number of important categories, including charge/weight ratio and a lack of "memory effect" - NiMH batteries need to be fully drained before recharging.
Despite these advantages, the dominance of Li-ion batteries in electric vehicles is not a foregone conclusion, as safety has always been a concern. A handful of recent incidents have highlighted this concern, including reports of a Model S catching fire and Boeing's Li-ion battery fires on the 787 Dreamliner. While safety concerns and competing battery technologies could pose a threat to lithium's demand outlook, we note that lithium has the highest charge/weight ratio of the main battery metals, including cobalt, manganese, nickel and phosphorus.
While electronics and other uses will continue to be important demand drivers for lithium, long-term projections for lithium demand hinge on the global penetration of electric vehicles, lithium's use in rechargeable vehicle batteries, and the mix of EVs, HEVs and PHEVs in the electric car market. In our view, the primary factor that will influence electric vehicle adoption is cost - of the cars themselves and of electricity - compared with traditional gas-powered vehicles. Further, electricity grid capacity and charging station infrastructure would also probably need to increase to support a widespread adoption of electric vehicles in the U.S.
In Anticipation of Strong Battery Demand, Lithium Capacity Has Already Grown Substantially
With further capacity additions planned, we think the market will probably be facing an oversupply situation toward the end of the decade. Lithium supply is fairly concentrated, with South America, China and the U.S. the main producing regions. Major lithium brine producers in South America include Rockwood (Chile), SQM (Chile) and FMC (NYSE:FMC) (Argentina). The majority of spodumene-based lithium production starts in Australia (Talison Lithium (OTC:TLTHF)), with final lithium carbonate production usually happening in China. China also produces lithium from brines, with salt lakes distributed across Qinghai, Tibet, Xinjiang and Inner Mongolia. However, based on the lower concentration and lack of scale, we think Chinese brine producers sit well above South American brine producers on the global cost curve.
There are numerous and sizable undeveloped lithium assets across the globe. The most notable is Uyuni in Bolivia, the largest known deposit in the world. This deposit is under consideration for large-scale development, but has a lithium concentration of only 0.05%, compared with 0.14% at Atacama. A trial plant in Bolivia will produce only 40 metric tons of lithium this year. The Bolivian government, which has kept a tight hold on the lithium resource, hopes to eventually produce 30,000 mtpa, which would be roughly 20% of 2012 demand. However, hurdles remain for the development of Bolivia's deposits, including the high amount of magnesium in the deposit (which drives up costs), a lack of freshwater and infrastructure and the reticence of the Bolivian government to use the help of an outside development partner.
While development of new deposits may face hurdles, current producers, especially brine producers, can expand production capacity relatively cheaply and quickly. Rockwood is wrapping up construction of a new 20,000 mtpa facility in Chile that will bring lithium carbonate equivalent capacity to 50,000 mtpa. And over the long run, we expect SQM to increase its production volume to 60,000 mtpa compared with about 45,000 mtpa in 2012. Another major South American brine producer, FMC, has capacity of about 23,000 mtpa, having increased capacity 30% over the past few years.
On the rock production side, a notable recent expansion project comes from Talison Lithium, in which Rockwood has agreed to purchase a 49% interest from Chinese company Tianqui, which outbid Rockwood for Talison. Talison is a spodumene-based producer operating out of the Greenbushes deposit in Western Australia. Talison produces lithium concentrate, a precursor to lithium carbonate, and exports most of its product for finishing in China. Talison recently completed a major expansion project, doubling capacity to enough lithium concentrate to make 100,000 mtpa LCE. It will be interesting to see how China makes use of this new capacity (and what sway Rockwood will have, given its interest), as Tianqui now has the ability to swamp the global lithium market. It's important to note that the Greenbushes mine likely has only about two decades of production left. This time frame could decrease considerably at higher production levels expected after expansion.
Other recent projects have been supported by offtake and supply agreements with battery makers. For example, Canada Lithium is ramping up production at a new mine with a capacity of 20,000 mtpa. We expect that other projects on the drawing board and in development will come to fruition. Considering this along with expansions from current producers, we expect oversupply in the lithium market in the years to come, despite our forecast for strong lithium demand. The exact level of oversupply is difficult to peg, given the relative ease with which capacity can be expanded at Atacama and other already-developed brine deposits. That said, we feel comfortable forecasting 2017 LCE supply of greater than 300,000 mtpa versus demand of about 260,000 mtpa.
We Expect Long-Term Price Pressure on Lithium
We use a $4,750/mt long-term price forecast for lithium carbonate in our assumptions that lead to narrow Morningstar Economic Moat Ratings for Rockwood Holdings and SQM. This forecast is in nominal 2017 dollars and implies a 2013 real price of about $4,300 in 2013 dollars. Our forecast is based on estimates of marginal costs of production and our view that recently added and upcoming additions to industry supply will probably suppress industry operating rates and prices. Our long-term forecast implies pressure on prices from the current level of above $5,000/mt. We admit that lithium supply and demand are both difficult to forecast. On the demand side, the penetration of electric vehicles, the mix of electric vehicles by type, and the penetration of lithium use in the car battery market are the primary drivers of variability. On the supply side, because of the scalability of lithium brine deposits in Chile, Rockwood and SQM can easily increase supply. With other lithium projects on the docket, Rockwood and SQM will need to decide how they want to shape the market. Even with strong demand growth, we think lithium suppliers should have sufficient capacity to meet demand, with the huge expansion by Talison the primary driver of a potentially loose lithium market to 2020.
Editor's Note: This article covers one or more stocks trading at less than $1 per share and/or with less than a $100 million market cap. Please be aware of the risks associated with these stocks.