They say history repeats itself, yet, somehow, this apparently doesn't hold true for lithium chemical supply. The choir singing the bearish anthem re supply estimates is growing, with Citi and Canaccord Genuity joining the ranks of Morgan Stanley in calling for massive oversupply until around 2024. These oversupply estimates reach as high as ~150K MT in 2023. Understandably, this would have serious repercussions for chemical prices, potentially pushing them as low as US$7,500/t. The market for lithium products is small but growing rapidly (20%+ p.a.), boom and bust cycles are inevitable. What history has shown us, is that substantially increasing the production of specialty lithium compounds, suitable for the applications where demand growth is most prevalent, is a very difficult task. The historical market for lithium has been mostly industrial applications and consumer electronics, the battery storage market (EV and ESS), responsible for 95% of the demand increase between 2018 and 2025, requires a much higher quality product.
The growth in lithium demand for these specific battery storage applications over the next 6 years is 8 times 2018 levels. What is most frustrating is that my 2025 demand and supply forecasts are not materially different from the analysts mentioned above, where we differ is in the timing of supply growth rates. This is a subtle but important difference, delaying supply growth by a year, given high demand growth rates, results in a far more balanced market.
Here are some of the reasons why I expect supply growth (of battery grade and battery quality chemicals) will be slower than anticipated:
- Incumbent brines producers are struggling to ramp up production. Through a combination of weather and water issues, technical challenges to improve product quality and royalty/transfer pricing disputes, 2019 supply increases from brine incumbents are likely to be modest. Additionally, SQM largely sells its output on a short-dated basis, without long-term contracts SQM is exposed to spot price volatility. The new CEO has stated that the company objective is to secure margins, not market share, intimating that they will build a strategic inventory. Based on early indicators SQM's production forecasts for 2019 are already under threat. Albemarle's Q1 2019 guidance, for both brine and hard rock output, suggests they too will struggle to meet 2019 targets. There are new brine projects and expansions coming online over the next few years underpinning some supply growth, however, the foundation of the oversupply thesis has shifted from brines to chemical production from spodumene concentrate.
- Spodumene concentrate producers are not delivering the quality and quantity of raw material as promised in their feasibility studies. The latest quarterly reports highlight (or try and avoid highlighting) the truth of the situation. It seems that weather can also adversely affect SC6.0 output, cyclone Veronica has been blamed for reduced shipment volumes in Q1 2019. Sub 6% SC is, in some instances, not being converted into battery-grade material. Even if battery-grade is achieved there is still the additional hurdle of qualification by ex China buyers that can take up to 12 months. For a more detailed discussion read Gerrit Fuelling's post-Not all Lithium is Equal and/or Lithium is Not a Commodity Recently Pilbara Minerals and Alliance Minerals have announced downstream partnerships. Pilbara has also delayed the completion of construction of stage 2 (to increase output to 800K-850K MT) to late 2020 in order to complete the search and appointment of a 20%-49% minority shareholder/partner. SC6.0 supply from stage 2 is unlikely to reach the market before 2021. Essentially SC producers are not making the margins suggested in their feasibility studies for a number of reasons (recovery rates, grades etc). It remains to be seen whether these issues will be long-term in nature. For some, I believe so. What is clear is SC producers have no incentive to risk compressing their margins further. Hedging their operating margins by gaining downstream chemical exposure is a logical step. Chinese converters have capital costs per ton of installed capacity of $6,000 or less. In order to achieve operating margins of 20% plus they can sell LiOH or Li2CO3 at prices well below their Australian competitors with higher capital (US$15k/t) and in some cases, operating costs. With SC6.0 contract pricing largely tied to China spot prices, SC producers will suffer as those prices fall. Below US$500/t some SC operations will become marginal - they need to avoid this outcome at all costs and delaying expansions and gaining some control over future chemical sales is a good survival strategy. In my opinion, there is a floor price for SC6.0, below which, assuming shareholders decide not to fund losses, supply will be removed from the market. IF existing and future SC operations achieved their production targets (volume and grade/impurities) on time then the oversupply thesis would have some merit. This is definitely not proving to be the case - expect further disappointments and delays as future quarterlies are released.
- Chinese conversion capacity - the scale of new chemical facilities are a measure bigger than historical plants. Whilst incumbents such as Tianqi and Ganfeng may achieve decent capacity utilization rates in time, newcomers are unlikely to enjoy similar successes. The quality and consistency of SC feedstock are critical, Tianqi and Albemarle are supplied by Greenbushes, the standout hard rock operation. Newcomers are being supplied by greenfield operations that have grade and impurity variations across their resources. Lack of processing skills is a further handicap to newcomers. As mentioned by Gerrit, even if newcomers achieve battery grade status, they still face an offshore quality qualification process of up to 12 months. In summary, the unprecedented production scale of new plants, variable feedstock quality and limited processing skills all point towards slower ramp-ups and delays in achieving export quality or simply battery grade chemicals.
- The high percentage of future supply derived from greenfield projects. Of the ~600K MT of new supply forecasted to come onstream by 2025, approximately 368K MT is greenfield in some way or form. Tianqi and Albemarle are both forecasting slow ramp-ups at their Kwinana and Kemerton plants. With Greenbushes as their feedstock and given their long histories of processing this material, these projects should be seen as the reference point for the rest of the industry, save for Ganfeng. Other projects will almost definitely face delays and possibly product disappointments.
- Capital costs and investment return hurdles. The average capital cost per ton of installed capacity, ex-China, is US$15,000+. As an extreme example, Albemarle's investment into Wodgina equates to US$40,000/t, assuming an 80% capacity utilization (generous given both the SC and chemical plant are greenfield), to achieve 2 times its weighted average cost of capital, Albemarle needs LiOH prices at US$14,000/t. If the market is oversupplied in 2022, would the JV sell hydroxide at US$8,000/t? It's extremely unlikely. Both Mineral Resources and Albemarle have the capital resources to delay production/sales from Wodgina. As it is, new projects are struggling to find backers at US$12,000/t+ prices. At US$8,000/t the capital markets for new projects will completely dry up, amplifying the supply shortfall in 2025. Supply growth from new projects is already under threat at current prices.
What are my chemical supply/demand estimates?
My model is forecasting very modest yearly oversupply between 2019 and 2023, in the order of 6K - 37K MT, split between hydroxide and carbonate. On balance, there is a greater oversupply of hydroxide during that period. These supply/demand forecasts can be broken down further into specific segments. Given the potential for future political issues, technical hurdles, poor weather and unplanned plant shutdowns, this oversupply could turn into a shortfall at any time. By 2024/2025 I expect material supply shortfalls. Further, as demand estimates are usage-based, any inventory builds at cathode or OEM (battery packs) level would add to current demand estimates.
At present, the analyst community is choosing to ignore the history of lithium supply growth and all the problems the industry has faced. The technical specifications of the vast majority of future chemical demand are more stringent than what's been produced historically. The list of hurdles faced by incumbents and newcomers is extensive. Based on the current trajectory of SC projects, brine projects and Chinese converters material oversupply is unlikely.
Demand for battery storage (EV and ESS) is likely to grow substantially and surprise on the upside given a) Chinese and European government regulations essentially forcing the adoption of EV's b) consumer preference for faster charging and longer range vehicles (increasing battery size) c) simple economics, BNEF research is forecasting large EV's in Europe to be cheaper than their ICE competitors by 2022.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.
Editor's Note: This article covers one or more microcap stocks. Please be aware of the risks associated with these stocks.