Evolving choices around EV battery composition have altered price dynamics in the lithium market, with the two main forms, hydroxide and carbonate, now moving independently to each other, reflecting different use cases and trading patterns.
Traditionally lithium hydroxide was produced from lithium carbonate, resulting in a high degree of price correlation between the two chemicals in the market. When Australia's spodumene capacity increased after new projects came online in 2018, the spodumene production route for lithium hydroxide became dominant. The increased Australian production, which altered supply routes, combined with a slower than anticipated growth in global EV sales and changes in Chinese policy, have combined to boost demand and liquidity in the lithium carbonate market, altering the price relationship between the products.
A key driver for lithium chemical demand and prices is the changing fortunes of different battery compositions. Lithium iron phosphate (LFP) chemistry has always been cheaper versus nickel manganese cobalt oxide (NCM) and lithium nickel cobalt aluminum oxide (NCA), which require expensive nickel and cobalt. LFP is also considered safer – as less prone to thermal runaways, thanks to the absence of nickel—but the trade-off has been lower range.
This last characteristic has been a sticking point: since one of the major barriers to mass adoption has been "range anxiety", LFP was considered by many industry participants to be a lower priority in the upcoming EV boom. It was frequently associated with low-end city cars in China, as well as electric buses or electric bikes and energy storage systems (ESS), but was seen as playing a minor role within the wider EV revolution.
This started to change with the gradual removal of Chinese EV subsidies, which lowered the incentives for local automakers to target only long-range EVs and increased the pressure to reduce costs.
The cost of an LFP battery is about Yuan 0.08/Wh, which is around Yuan0.15/Wh-Yuan0.21/Wh cheaper compared to ternary cathode materials (NCM), corresponding to a cost reduction of 65%-72%, according to ICC Sino.
Moreover, improvements on the pack design, through the cell-to-pack approach, allowed a bigger portion of the battery pack to be filled with cells which significantly increased energy density. Tesla's Model 3 Standard Range produced in the Shanghai factory featuring LFP batteries supplied by CATL have around 450 km driving range; BYD's Han model, using the so-called Blade Battery (LFP, cell-to-pack) reportedly has a 605 km range.
These are comparable to the driving ranges of cars with NCM 5:2:3 batteries at a much smaller cost. BYD, which is China's largest seller of electric vehicles and ranks only behind of Tesla globally, recently announced it will scrap its NCM technology and employ only LFP going forward.
Characteristics and contracts
Although many agree that hydroxide will be a major lithium chemical in the future, there is a consensus that carbonate still represents the majority of the market at the moment. In addition to its predominance in volume, carbonate is also easier to handle and has a longer shelf life. These characteristics allow carbonate to be more frequently traded in the spot market than hydroxide.
The biggest hydroxide buyers are battery makers in Japan and South Korea who are inherently more prone to signing long-term contracts because of the nature of hydroxide and the necessary focus on specifications.
In many cases, these companies supply nickel-rich material to Western original equipment manufacturers (OEMs) producing high-end EVs. This means, firstly, that the required logistics are complex, since lithium hydroxide must be shipped from China (or alternatively from farther locations such as Russia or the Americas) to Japan/South Korea to be used in EVs ultimately sold in the West. This process requires very stringent planning, and consequently the spot market has typically been used only as a last resort.
Secondly, quality requirements will always be very strict. As already mentioned, carbonate does not currently suit nickel-rich chemistries due to temperature requirements—nickel-rich chemistries are more complex and possess some safety concerns. Raw materials specifications are therefore paramount when dealing with nickel-rich chemistries, and it has been desirable for lithium hydroxide users to develop long-term contracts which have typically been on a fixed price basis, and so not necessarily reflective of the market value over the given period. As most of these major hydroxide buyers are in Japan and South Korea, price stability and predictability have been more desired, and therefore long-term (which can often mean multi-year) contracts under fixed pricing has been the buyers' preference.
Carbonate and hydroxide prices diverge
Before the shifts in the battery metals and the EV market described above, lithium hydroxide, as a product derived from carbonate, was always priced based on the presiding carbonate price plus the additional processing cost, so both prices largely moved in tandem.
The processing cost was estimated to be around $1,500-$2,000/mt, and for some time this represented the typical carbonate-hydroxide spread. In 2019-2020, when lithium spot prices were moving down consistently due to rising Australian spodumene production, combined with the slowdown in global EV sales, the carbonate-hydroxide spread climbed to as high as $4,000/mt in the international market (S&P Global Platts CIF North Asia prices) in March 2019.
One of the two main reasons for this was the lower liquidity in the lithium hydroxide spot market compared to lithium carbonate, meaning that hydroxide spot prices are more static due to long-term existing contracts. The second was the widespread belief at that time, despite some persisting difficulties and safety issues, that high-nickel chemistries (mainly NCM 8:1:1) will dominate the market in the future due to its inherent higher energy density, vis-à-vis superior range.
As shown in the charts, the carbonate-hydroxide spread in the international market has fluctuated in the last two years and then flattened out at $1,000/mt from late January to early April 2021. In contrast, in the Chinese domestic market (Platts DDP China prices), the carbonate-hydroxide spread shrunk at a far quicker pace at the start of the period, and by the end of 2019 the spread had thinned to less than $400/mt—it never reached spreads as wide as those seen on the international market at their peak. Carbonate moved up rapidly in China in late 2020, reaching parity with hydroxide at the end of the year. In 2021 carbonate has strengthened further, hitting a $2,500/mt premium to the hydroxide price by March.
In both the international and Chinese markets there was a strengthening of carbonate versus hydroxide, either reducing the differential to historical lows or inverting the previous price lead. The main drivers that led to this new scenario were the resurgence of LFP in the Chinese market, and lithium hydroxide's smaller spot market compared to the liquid carbonate market.
In China, where the lithium spot market is much more active than in other regions, the share of hydroxide volume traded under long-term contracts is much bigger than that of carbonate. The limited liquidity on the hydroxide spot market, as a result of the predominance of long-term contracts in this segment, reduced volatility. This was observed during the 2018-2020 lithium bear run, when carbonate prices fell faster than those of hydroxide, and is being proved again in the bull run, with carbonate moving up more rapidly too.
Nevertheless, as nickel-rich chemistries grow in importance, given the expectations of adoption across Europe and the Americas, demand for hydroxide will certainly increase. This will undoubtedly bring with it price instability, given the inherent risk of demand outpacing supply in years to come, but to a far lesser extent than what will likely be seen in the carbonate markets.
Chemistries in the future
The LFP resurgence in China was one of the key factors driving lithium carbonate prices above lithium hydroxide, but with the global EV axis progressively moving to Europe – the region is expected to become the biggest market by the end-2021—the LFP story was expected by many to fade out.
The European market is still expected by several industry participants to be largely dominated by nickel-rich batteries such as the NCM 8:1:1, which provides higher energy density. The typical European consumer, being concerned with the range of an EV, would be expected to pay a higher price tag than the typical Chinese consumer. However, promoters of the nickel-rich battery are potentially overestimating a typical commute in Europe (especially when compared to the US) and overstating the range anxiety among European consumers.
Nevertheless, if a longer-range EV is on offer at a competitive price point, it will likely be chosen by the majority of consumers despite the little need for long range. It was on this assumption that most of the investments in battery plants in Europe were targeted to nickel-rich chemistries.
This makes even more sense at the moment, since Europe has just started to provide subsidies for EVs, particularly in Germany and France. However, it remains to be seen whether the logic continues to hold when subsidies start to be removed and cost becomes increasingly important to the general consumer.
It is important to emphasize that battery-grade nickel and cobalt chemicals are moving towards a significant deficit in the next few years, in a similar pattern to lithium. This would add to the prevailing cost advantage of LFP versus nickel-rich chemistries.
Taking the lower cost into consideration, combined with the safety factors and technology improvements that enhanced energy density, the LFP resurgence in China was not entirely unexpected and the connection with the subsidies phaseout is clear. Still, China and Europe are each at different stages of their EV experience.
China was an "early adopter": it offered subsidies since 2010, and is now in the latest stage of removing incentives to let the market grow organically, while Europe has just recently adopted the support strategy. Therefore, there will be significant growth of NCM 8:1:1 powered EVs made in Europe over the next few years, yet a significant portion of the market for short-to-medium range EVs powered will likely be powered by LFP and other non-nickel-rich alternatives. Similar to the Chinese experience, these subsidies will not last in the longer term, meaning the carbonate-based chemistries share could increase at a later stage when cost becomes a key driver.
Evidencing this, Volkswagen recently announced it will use LFP in its entry models, as well as a manganese-rich chemistry for volume models, and leave nickel-rich NCM for "specific solutions." Volkswagen's top selling car is the VW Golf, a volume model that will probably not feature a nickel-rich cathode once electrified. Tesla's CEO, Elon Musk, also said the US EV maker could expand the use of LFP too. Even if hydroxide demand grows faster than that for carbonate, it seems increasingly unlikely that hydroxide will account for 60-70% of the demand, as some market participants once believed.
Technology will also play a role in further improving energy density, regardless of the cathode chemistry choice. Several companies are working to increase the proportion of silicon in the anode, as well as in developing solid state batteries.
Novel LFP-based batteries can reach around 200 Wh/kg and NCM 8:1:1 provides 300 Wh/kg, while it is believed that a solid state battery can reach 500 Wh/kg, meaning that by the end of this decade, the cathode chemistry choice is likely to be less of a determinant than presently when considering energy density. There are also other innovations being tested, such as from technology company Nano One, that could allow the production of nickel-rich cathode materials using carbonate instead hydroxide.
Demand for hydroxide could possibly grow at a quicker pace than for carbonate in a subsidized European EV market, where nickel-rich chemistries are expected to be mainstream—although it is still unclear how competitors will react to Volkswagen prioritizing cost as mentioned. Nevertheless, an "LFP resurgence", this time in Europe, cannot be ruled out in the future, especially as OEMs need to increase price competitiveness, which will be exacerbated when subsidies start to fade away.
In the case of the Chinese domestic market, cost and consumer preference are indicating continued utilization of LFP—and consequently, carbonate— for the foreseeable future. Although carbonate and hydroxide prices might converge and reduce the current carbonate premium, there are no apparent reasons for hydroxide prices to go back above carbonate in China anytime soon.
Beyond 2021, the use of nickel-rich chemistries that require hydroxide is also expected to grow in China, but non-nickel rich chemistries are likely to keep accounting for the biggest portion of the market. Included in this equation is the fact that China has substantial conversion capacity, allowing for either carbonate or hydroxide production from spodumene at virtually the same cost, therefore it seems likely that carbonate and hydroxide prices will tend to converge over time, with thin spreads and temporary price advantages for each product depending on consumer choice over time.
Lithium carbonate will continue to account for a significant portion of the market throughout this decade and will play a critical role in the EV revolution. This is not necessarily the case for what will be the top-end models, but with several OEMs aiming to become fully electric in the next few years, it's reasonable to think that most of the newly electrified cars will be a range of models catering for different needs.
Lithium hydroxide's price edge over carbonate in the seaborne market is more likely to stay in touching distance to current levels than return to the vast gap seen through 2019-2020. In the Chinese market, carbonate prices should keep trading above hydroxide in the short term, and move to be mostly on parity in the longer term given China's dominant conversion capability.