Thematic, innovative and bespoke investment solutions and thinking

# The Bear Case on Lithium –

With climate change top of mind, investors are fully aware that battery metals will play a crucial role in the 21st century […].  Yet despite this exponential demand profile, we see the battery metals bull market as over for now.

Goldman Sachs Global Investment Research, 29 May 2022.

# Negative on battery metals? Really?

I remain convinced that the long-term outlook for virtually all green metals is rock-solid.  Having said that, an outsized supply response to substantial price increases on most battery metals has forced me to sit up and listen carefully to the bearish camp – in this case represented by the commodities team at Goldman Sachs (GS).

We publish investment strategies and opportunities in our research papers. This research paper is available to professional investors as part of ARP+ subscription.

In short, they have concluded that three key battery metals – cobalt, lithium and nickel – have either peaked already or are about to peak price-wise, at least temporarily.  Oversupply will put those three metals on a downward trajectory, GS believe, which should last 2-3 years.  In terms of price impact, the research team at GS project that lithium will be affected the most, followed by cobalt.  The impact on nickel will be less dramatic but still meaningful.  GS’s model suggests that, between 2022 and 2025, lithium supply will grow by 33% annually, cobalt by 14% annually and nickel by 8% annually.  Those supply forecasts should be held up against expected annualised demand growth of 27%, 11% and 7% respectively.

In other words, according to GS, in all three cases, supply growth will outstrip demand growth over the next few years.  Now, although I remain long-term bullish, I simply cannot ignore such a view.  My objective with this paper is therefore to dig deeper on the views expressed by GS.  In the following, in order to avoid any confusion, I will make it as clear as I possibly can whether the views expressed are those of others or my own.

# Demand and supply

According to GS, in order to achieve the Net-Zero goals already established, road emissions must be reduced by at least 13% by 2030.  This implies that, by 2030, at least 70% of all private vehicle sales must be electric.  At the same time, there is an urgent need to dramatically improve the ability to store electricity, as you are prone to regular overproduction when most electricity comes from renewable energy forms – for example during times of high winds.  The research team at GS estimate that the combination of those two factors – high EV demand and high grid battery demand – will translate into a rise in total battery demand of no less than 440% between now and 2030.

Demand for cobalt and nickel should both grow in line with that number.  Demand for lithium should grow even faster, as this metal is also used extensively in other industries.  In the medical industry, it is used to treat for example mental diseases like bipolar disorder, depression and schizophrenia.  It is also used to treat alcoholism, diabetes and arthritis.  The glass and ceramics industry is another significant user, and so is the aircraft manufacturing industry.  Having said that, in the years to come, the primary growth engine will almost certainly be the battery industry, which is expected to account for no less than 95% of all lithium consumption by 2030 (Exhibit 1).

As far as supply is concerned, as per McKinsey & Company’s (McK’s) latest count, virtually all lithium – 98% to be precise – comes from either Latin America, Oceania (almost exclusively Australia) or China – see Exhibit 2.  As per McK, over the next few years, a number of new production sites will pop up all over the world, but that shall also be required to meet the expected growth in demand.  As you can see below, McK do not expect supply to grow faster than 20% per annum between now and 2030, which is considerably less than the 33% annual growth rate projected by GS.  In other words, two of the leading firms in our industry have vastly different supply expectations.  More on that later.

# The Goldman Sachs bear story

Cobalt, lithium and nickel prices have all benefitted from demand outstripping supply in recent years.  As you can see in Exhibit 3 below, in 2021, both cobalt and lithium were severely undersupplied (10%+ each) – nickel less so, but demand still outstripped supply by about 6% according to GS.

However, as you can also see, GS expect supply to quickly catch up with demand with the two being well-balanced this year.  As early as next year, GS believe lithium supply to dramatically outstrip demand – a trend which will further intensify in 2024.  New projects in China, Australia and Chile are responsible for the projected ramp-up in lithium supply, they say.

Before going any further, I need to make a clarifying point.  As such, there is no single lithium product or price.  The lithium market is very complex with a myriad of grades and prices.  At the highest level, lithium comes as lithium carbonate or lithium hydroxide, and there are several grades underneath those two main types of lithium.  Prices vary considerably across those different grades.

A good example is battery-grade lithium.  Whether the lithium is battery-grade or not will have a significant impact on the price.  For instance, a difference of 10 parts per million either way can make a massive difference.  If your lithium carbonate has 20ppm of iron, you may not be able to use it for batteries.  If it has 10ppm, then you probably can.  Consequently,  one has to be slightly cautious when comparing different estimates.

According to Trading Economics, lithium carbonate currently trades at CNY 477,500 or about $71,000 per tonne. The commodities team at GS expect that price to average less than$54,000/tonne this year and only $16-17,000/tonne next year, i.e. a steep price drop is around the corner if you believe them. Cobalt should be less affected and nickel even less so. GS expect the cobalt price to average$59,500/tonne next year vs. a current spot price of about $71,500/tonne. The story on nickel is somewhat different. A short-term supply crunch should result in higher prices over the next 5-6 months, following which the same oversupply dynamics should kick in and drive the price on nickel lower. As per GS, this should result in nickel appreciating from about$23,000/tonne at present to $36,500/tonne by year-end, before the price begins to drop. GS expect the nickel price to average about$30,000/tonne next year.  If you take a look at Exhibit 4, you can see that the lithium price should be affected far more dramatically than the other two battery metals, should GS’s projections prove correct.

# Why Goldman Sachs may not be correct

The first voices of dissent have already surfaced.  In a note dated the 1st of June, i.e. only three days after the GS research paper was published, Matt Fernley of Westbeck Capital Mgmt. argued that what in theory is quite simple (adding new capacity) is, in practice, a great deal more complex.

He is of the opinion that many sell-side firms underestimate the time it takes to add new lithium supplies but did not mention GS specifically in his note.  A brine project takes at least 2-3 years to get up and running, he argued.  He said that it is impossible to speed up the process, as it takes at least 12-18 months, and in some countries considerably longer, for the water to evaporate from the brine pools.  He added that it is even worse if the extraction is based on rock mining.  Some sell-side firms claim that a hard rock value chain can be in production within two years, but such an argument is deeply flawed, he said.

Apparently, the commodities team at GS have only published research on lithium for a couple of years (whether they have followed it for longer, I don’t know), and I have been informed that the analysts involved are raw materials generalists with only limited coverage of the lithium universe.  It is also worth noting that the recent GS paper did not represent a change of opinion on lithium.  In fact, they have been bearish on lithium since they first picked up coverage a couple of years ago.

Another ardent protest was filed only a few days later.  In a post on its website (which you can find here), Benchmark Mineral Intelligence (BMI) argued that there are no less than five reasons why GS’s analysis is wrong.  I shall not go through those five reasons here, but they are well argued and deserve ten minutes of your time.  McK have also been very active on the topic more recently with their views being more aligned with those of BMI and Matt Fernley than those of GS.

I cannot comment on the quality and experience of the McK team, as I don’t know them, but I happen to know how experienced Matt Fernley and the BMI team are.  In both of those cases, we are talking 20+ years of experience.  Matt Fearnley tells me that he has reached out to a number of sector specialists in China, none of whom believe the new production sites can be up and running as quickly as GS believe.  In that context, it is also worth mentioning that GS made no effort in its recent research paper to distinguish between battery and non-battery grade in the forthcoming supply boost.

# Brine vs. rock mining

A good example of the limited amount of details in the GS research paper is the no-mention of brine vs. rock mining in the extraction process; however, it is important to understand how the two extraction techniques vary, and how that may impact our investment strategy.  Let me explain.

Lithium is a soft, white-ish metal that almost looks like silver. It is very light – it is actually the lightest of all alkali metals.  It is either mined from solid rock, or it is extracted from underground pools of brine.  The two extraction techniques are very different, and one is significantly more expensive than the other, so it is important to understand the difference between the two methodologies.

Brine first:  When deploying this technique, lithium is extracted from brine by pumping brine into large evaporation pools, which have been established for the purpose (Exhibit 5).  The end-product, when deploying this technique, is lithium carbonate and is produced mostly in South America.

Hard rock mining of lithium, on the other hand, is not dissimilar to other types of hard rock mining.  Mining companies that deploy the hard rock technique like to argue that they can produce lithium cheaper than brine producers can, but that is a flawed argument.  Once you include the entire process – from prospecting and discovery to development of the site, exploration and mining plus transport from site to nearest transport centre (e.g. harbour), brine production is much cheaper.  There is evidence of that if you compare operating margins of two companies deploying different techniques.  (A good comparison can be found here.)  It is also worth noting that lithium from brine is a great deal less environmentally damaging.  On the flip side, the processing time is significantly longer, as it takes a considerable amount of time for the water to evaporate.

The end-product, when extracting lithium from rock mining, is either lithium hydroxide or lithium carbonate, depending on the processing technique deployed.  Operating margins when extracting lithium from hard rock mining are, on average, only about half the margins brine producers enjoy (source: S&P Global).

As you can see in Exhibit 6, brine-producing countries like Chile and Argentina can produce lithium at a fraction of the cost of those countries that extract most of their lithium from hard rock.  That said, even if the cost advantage of brine relative to rock mining should lead the industry to gravitate towards brine over time, the limited number of brine pools available will almost certainly keep the lithium mining industry alive for many years to come.

By far the biggest concentration of brine pools anywhere in the world is in the so-called lithium triangle – an area in South America bordering Bolivia, Chile and Argentina.  The ample presence of brine in the lithium triangle combined with world-beating production costs have made Chile the epicentre of lithium extraction (Exhibit 7).  Precise numbers as to how total lithium production is split between brines and rock mining do not exist, but industry insiders have told me it is about 50:50.

One point worth bringing up in this context is Bolivia.  As is evident when looking at Exhibit 7, hardly anything comes out of this country at present, and the same is the case if one looks at which countries sit on the biggest lithium reserves (Exhibit 8).  As you can see, Bolivia’s reserves are apparently too small to even get a mention.  However, industry insiders tell me that the world’s largest lithium reserves are to be found in the Bolivian part of the lithium triangle.  Apparently, the country is behind in terms of getting organised and taking advantage of its vast reserves.  In other words, a supply shock may be forthcoming at some future point, should the Bolivians sort themselves out.

Having said that, Matt Fernley also tells me that one of the reasons the Bolivian brines haven’t been developed yet is that they’re rich in magnesium.  High magnesium levels in brine impact the separation process and lower the lithium recovery process.  For the same reason, Argentina may not be able to fully develop its considerable brine resources.

# Investment implications

If – and that is a big if – the lithium price is going to drop as precipitously as forecast by GS, the negative impact on high-cost producers will be much bigger than it will be on low-cost producers, most of whom operate in the lithium triangle in South America.  On the other hand, if GS’ critics are correct that it will take much longer for supply to catch up with demand, the lithium price will probably recover quite quickly from the GS-induced losses and then continue to appreciate.  Such an outcome should favour high-cost producers.  Thus, the two outcomes warrant two very different investment strategies.  Herein lies the dilemma.

A somewhat similar scenario unfolded in 2018-19.  Demand had grown quite rapidly in 2015-17, leading many new projects to be initiated and, suddenly, global markets were flooded with lithium.  As a result, as you can see in Exhibit 9 below, the lithium price dropped almost 50% in 2018-19.  What you cannot see, but what I noticed back then, was that low-cost producers weathered the storm much better than high-cost producers did.  That is why I would favour a long/short strategy – long low-cost and short high-cost producers – if I knew for certain that GS were correct.

The long portfolio should probably consist mostly of companies operating in the lithium triangle, and Chile would be my favourite hunting ground.  Most lithium mining companies have sold off since GS published its research paper in late May.  Take for example the Chilean blue chip company, Sociedad Quimica y Minera de Chile (SQM on NYSE), which closed at $113.33 the day before GS published its research paper in late May. As of Wednesday of this week, SQM traded at$86.97 – off about 23%.

Another attractive investment opportunity is a smaller company called Lithium Chile.  I recently met with the senior management team of the company and believe it offers a desirable entry point at current price levels.  Chengxin Lithium Group Company have recently acquired 19.86% of Lithium Chile for $34Mn – a clear indication that Chinese companies are raising their stakes in this area. Is Lithium Chile in play? Only time can tell, but the combination of vast reserves, very competitive production costs and a cash-rich part-owner, who has taken a very active interest in further developing Lithium Chile’s reserves, can only be good news. Lithium Chile is a penny stock, though, and traded at C$0.61 on Wednesday.  The obvious implication is that you should treat it as an option, i.e. allocate much less to it than you would to for example SQM.

If, on the other hand, GS is proven wrong, high-cost producers will most likely outperform low-cost producers, even if they all do relatively well.  Given that I have no unique insight into whether GS is right or wrong, and given how much all lithium mining companies have sold off in recent weeks, I would recommend the following strategy:

1. For now, only establish a long portfolio consisting of low-cost producers. If GS is proven wrong, they will probably not perform as well as high-cost producers will, but they offer a much better hedge, should GS be correct.

2. If the evidence is growing that GS could be (more or less) correct, keep the long portfolio of low-cost producers but add a short portfolio of high-cost producers (on an equity beta-adjusted basis).

If going short is not an option you have, I would suggest you instead go long a basket of the lowest-cost producers you can find, and then gradually convert to higher-cost producers, as we approach the end of the storm.  Why?  Because high-cost producers will most likely begin to outperform low-cost producers as we come out on the other side.

# Final few words

I have made no mention of the ongoing war in Ukraine so far, but that could also affect demand for lithium over the next few years.  The longer that war drags out, the more likely it is that GS will be proven wrong.

Several (mostly European) countries are now planning to intensify the green transition to make themselves less dependent on fossil fuels from Russia.  That said, I came across an article in the Financial Times the other day, which suggested that Germany, Austria and the Netherlands are all looking to re-start decommissioned coal-fired power plants, as Russia cuts back on gas deliveries.  You can read the story here.  Precisely how this will all pan out is difficult to say at this stage, though.

Niels C. Jensen

29 June 2022