Posted on - 27 Dec 2021
It’s the question on everyone’s lips and we had a crack at answering it in the December issue of Battery Materials Review. It’s a difficult question because lithium prices are now in virgin territory (they’ve never been this high) and how do you model something when you don’t have a frame of reference?
Well, you have to look at other materials to be your guide and that’s what we used in our approach. To cut to the chase, we suggest that lithium prices, which are already up 500% from the Q3/20 trough for SpodCon and 700% for BG lithium carbonate, could rise as high as 1300% from the trough.
So, to put that in plain numbers, that’s SpodCon prices of over US$5000/t and BG lithium carbonate prices of over US$60/kg in this cycle. That’s compared to SpodCon prices now of US$2360/t, China spot BG LC prices of US$38.8/kg and contract prices around US$20-25/kg. So, that’s still quite a big increase from here, for all those cautioning that the lithium event is over.
I’m not going to go into my methodology here. Feel free to subscribe to Battery Materials Review if you’d like to learn more about my modelling approach and its risks.
Maybe some readers think I’m nuts! Many people thought I was nuts as well when I forecast in August 2020 that lithium prices could increase four times from their prevailing levels. As it turned out, I was too conservative that time. Maybe I’ll be too conservative this time as well?
While we’re talking about overly-conservative forecasts, I’d like to talk about what the sell-side are using in their models for stocks. I was updating my consensus forecast data this week and it makes interesting reading.
It’s probably not a surprise that sell-side consensus is forecasting lithium prices to increase next year. After all, most analysts (including us) are forecasting a significant lithium deficit. But what may be a surprise is that most sell-side houses are forecasting 2022 lithium prices below current spot! Yes, you read that correctly. Sell-side equity analyst consensus has SC6 prices averaging US$1861/t in 2022, 21% below current spot, and BG LC prices at US$25.3/kg, 35% below current spot.
What does that mean? It means that equity analysts are massively out of the money on their price forecasts and that stock valuations are going to have to increase substantially from here. Just to give you an idea of how far analysts are out – naming no names, but a major US bank which famously downgraded the lithium sector earlier this month is utilising a contract LC price of US$12.70/kg for 2022E, which is almost 50% below the Q4/21 contract price for LC!! There is only one broker on our list that has 2022 price forecasts in line with or above current spot prices. All the rest are below, many markedly so.
So, in a situation where the lithium market is going to stay in deficit, most sell-side houses seem to be implying that lithium prices will come down. Of course, they won’t if the market remains in deficit, so analysts will need to increase their price forecasts (likely substantially), which will have a knock-on effect on equity valuations.
To be fair, to some extent the market is already looking beyond the sell-side’s forecasts. That’s why stocks have overshot existing analysts’ price targets and we’re starting to see broker downgrades. But in this environment, I’m afraid I believe that it’s analysts that are wrong, not the market!
A recurring theme in the lithium industry this year has been the difference between the Chinese spot price for lithium chemicals and the longer-term contract pricing level. Most of the established lithium producers started the year selling into annual price contracts, but that structure is likely to change from here. At this point, the difference between spot and contract lithium prices is huge; in carbonate it could be the difference between US$38/kg for spot material in China and US$7/kg for material delivered to Korea on annual contracts.
Talking to traders and market participants though, this is changing rapidly. As we saw with iron ore in the 2000s, annual pricing contracts are now changing to quarterly pricing contracts. And those quarterly contracts have tighter links to spot pricing. Talking to traders, they suggest that more recent contracts are being linked to Asian spot lithium prices, sometimes Chinese spot prices.
Which means that there is substantial potential for companies with a lot of exposure to annual price contracts to see a material uplift in realised revenues in Q1/22. Indeed, some are seeing it already. At its Q3/21 results, SQM noted that it expected selling prices to increase 50% in the fourth quarter. On our Recharge podcast this month, Ken Brinsden, CEO of Pilbara Minerals, talked about renegotiating longer term contracts with customers so that they are closer to spot pricing. He mentioned that his company’s contract prices were now c.US$1650-1800/t for CQ4/21 from under US$1000/t previously. I would suggest that they’re likely to go higher still in 2022.
This move from annual to quarterly pricing means that even those companies selling on long-term volume contracts will be able to take the benefit of increases in spot prices. The delay in pass through is shorter too. At the beginning of this year there was a 4-5 month time lag for price pass through. Now it’s closer to 1-2 months. This change in price realisations is not really being reflected by many sell-side analysts currently.
This, of course, is the next big question in the sector. When we think back to the commodity supercycle in the 2000s, the price increase wasn’t a straight line event, and there were periods when prices fell back, as well as increased.
These fall-back periods were generally the result of one of two key drivers:
1) Perception of changed macro conditions for demand; and
2) Short term supply/demand imbalance.
Now, for me, there is quite a substantial difference between the price behaviour of exchange-traded commodities like copper, zinc and crude oil compared to lithium. I don’t expect lithium prices to be so elastic to overall economic cycles because the major driver of lithium demand (the EV event) isn’t. So I don’t see driver (1) as such a key driver for lithium prices.
I do see driver (2) as an important one for lithium, however. And I think in this context any slowdown in the rate of EV sales and/or start-up (restart) of large projects could create short-term supply/demand imbalances in the sector.
While we’re seeing ongoing expansion of capacity in the Atacama, for me it’s going to be hard rock start-ups in Western Australia and elsewhere which have the potential to destabilise the market in the next 12-18 months. However, due to Covid delays, I don’t expect any material impacts until the end of 2022 at the earliest, most likely early 2023.
That means I’m pretty happy that the lithium market will stay in deficit for most of 2022.
Beyond that there is a fair amount of capacity that will start to come on, but I think most of it is going to be offset by higher demand. I expect lithium prices to be a bit more cyclical than they have been so far, but I would still expect them to be on a generally rising trend over the next several years.
The reason being that the thing that’s holding back our EV sales forecasts currently is a likely shortage of lithium. As lithium comes into the market, more batteries can be made and more EVs can be sold. I don’t expect a long-term imbalance to occur at any time during this decade.
I know a number of sell-side analysts are not as positive as myself and many of the other industry specialists, but I would suggest that some non-specialist analysts are getting technical issues wrong in their supply/demand modelling. The most significant errors that I see in models are:
• The assumption that 100% of material being produced from brine in Latin America is battery grade (when, in fact, it’s closer to 80%). While a lot of this non-BG material can be upgraded, there is a recovery loss of c.7-15% on this. If analysts are not factoring in this loss, then they are overstating supply.
• The assumption that there is a 100% recovery on hard rock material being converted in China (or elsewhere). Again, average recovery levels for this are 85-90% of lithium units (and much lower at the beginning of a converter’s life). If analysts are not factoring in this loss, then they are overstating supply.
• The assumption that all new projects will come into production on time and on spec. There hasn’t been a new project yet that has come into production on time and on spec. Most projects take at least 6-12 months to ramp up. If analysts are not factoring this in, then they are overstating near-term supply.
• A dirty secret of the battery industry is the issue of battery yield. Battery yield describes the share of faulty batteries that are produced. So, if you have a battery factory with nominal capacity of 10GWh, then it will likely only produce 8GWh of functional batteries, the rest being waste which needs to be recycled (and therefore may not be available as battery raw materials for some time). Most analysts forecast in terms of EV sales, so unless they factor in battery yield (which can be much lower in the first years of operation), they are hugely understating demand for battery raw materials.
A notable sell-side house which has a very bearish view on the lithium industry doesn’t seem to have included any of these factors in its supply/demand model, which is calling for huge oversupply and lithium prices to go below what I regard as the marginal cost of production from 2023 onwards. Needless to say, I don’t regard their research as top quality!!
But to all seasons, there must come an end. And the end will come for the lithium price event, be that in the second half of this decade, or in the early 2030s. It will come, as with any raw material price cycle, because there will be an overinvestment in new supply. And, when that new supply comes into the market, prices will fall.
But how low will lithium prices fall? Well, I would suggest that short-term corrections are likely to be of the order of 20-40% but when the collapse comes, it could be much more significant than that.
My view is that prices could very well fall to the marginal cost of production. But how does one define the marginal cost of production? In a mature market like copper, I would suggest that the marginal cost is around the 66th percentile of production costs (so that’s operating costs, not all-in costs, but similarly not cash costs either). But in a market that’s relying on expansion to feed secular demand, then you need to factor in some element of capital cost as well.
And, additional to that, while the current marginal cost of production for lithium carbonate is likely a factor of the operating costs of SQM’s and Albemarle’s Atacama operations, as the industry grows and there are more operations, it won’t be. The third and fourth generation of lithium operations will have higher operating costs and that will push the industry’s overall marginal cost upwards.
Added to which – I’m very much of the view that we’re moving into a high inflation decade, so labour costs, energy costs, transportation costs, mining costs will all increase over the next 5-10 years. That’s going to drive marginal costs up as well.
I’ve written about this a number of times in BMR so won’t labour the point here, but it’s my view that the marginal cost of production of lithium chemicals by the mid- to late-2020s will be 40-50% higher than the marginal cost was in 2019-20. So I don’t believe that lithium prices will ever go back to the 2020 trough levels.
Matt Fernley is Editor of Battery Materials Review and Head of Research for Westbeck Capital’s Volta Energy Transition fund.