Lithium - Where to From Here?

Airlie Funds Management

17 January 2022

Following on from our report last month on Mineral Resources and in particular its lithium business, we thought our background work on the lithium industry and supply/demand fundamentals might be of interest:

Over the past 12 months, the share price movements for ASX-listed lithium producers and developers have been eye-watering.  Pure-play producer Pilbara Minerals (PLS) is up +280%.  Nickel / lithium producer IGO is up +80%, having completely reshaped their entire business through a procession of transactions to become an integrated "battery metals" business.  Iron ore / lithium / mining services company Mineral Resources (MIN - an Airlie favourite) is up +50%, despite a 27% decline in the iron ore price over the same time.  Listed developers / explorers Liontown Resources (LTR), Firefinch (FFX) and Core Lithium (CXO) are up +318%, +380% and +270% respectively (!).  The sceptic in me is wary of being the greater fool in the lithium sector, but pragmatically the equation is simple - even the most conservative estimates see the lithium market entering a material deficit at some point over the next 10 years as the electrification of transport takes place globally. 

Figure 1 - Airlie Funds Management

The potential for a material lithium supply deficit means prices for lithium raw materials and chemical products could continue to rise.  As more lithium extraction volumes and processing capacity comes online, the cost curve for lithium will continue to evolve and potentially result in a range of economic outcomes for companies depending on their cost position and capital invested.  Herein lies the opportunity and uncertainty for investing in lithium extractors and processors going forward.

Demand

Lithium mining and processing aren't new industries, but they are experiencing a structural change in their demand profile.  Lithium is a chemical element that doesn't occur freely in nature, but only in compounds, and is generally extracted from hard-rock despots or brines and then processed into a useable chemical product.  Historically, demand for lithium chemical products has come from applications in glass and ceramics, as well as additives in steel and aluminium production.  Today, due to the superior energy-to-weight characteristics of lithium, lithium chemical products have become an important component of the rechargeable battery cells that can be found in most modern electric vehicles.  As the world looks to transition away from fossil fuels, the demand for electric vehicles, and subsequently lithium chemical products, is robust.

Figure 2 - Albemarle Investor Day September 2021; Roskill 

The trouble is, while lithium is not exactly scare, the supply chain from raw material to useable chemical product is still developing as demand grows rapidly. 

Figure 3 - Lithium Battery Supply Chain, Deutsche Bank Lithium 101 May 2016 

Extraction

As mentioned earlier, lithium must be extracted (via hard-rock mines or brine lakes) and then processed into a useable chemical product.  The Australian lithium extraction industry is dominated by hard-rock assets (mines) that produce an ore which contains the lithium-bearing mineral spodumene.  Like all ore bodies, hard rock spodumene deposits can vary in size and grade, which ultimately affects the quantity, quality and cost of the product produced (see Australian spodumene cost curve below).  Spodumene must be processed into a concentrate of a suitable grade before it can be processed into a chemical product, and thus higher-grade ore bodies can have significantly less costly pathways to final product.  Brine assets (most found in South America) take saline brines with high lithium content and pump them from below the earth's surface into a series of evaporation ponds from which a more concentrated lithium-brine is produced. 

Figure 4 - BofA Global Research, January 2022

Processing

Spodumene concentrate can be converted directly to lithium hydroxide, while brine assets ultimately produce a lithium carbonate, which can then be further treated to create a hydroxide product if necessary.  To further complicate things, not all processing assets are integrated with upstream raw material extraction assets, meaning they must procure raw materials (i.e., spodumene or lithium carbonate) from producers.  Currently China has the dominant share of downstream lithium conversion capacity, a function of a historical cost advantage and proximity to customers.  Australia's share of global downstream conversion is significantly smaller than its extracted share and will remain so even as assets currently under construction come online.  Increasingly, Australian spodumene producers are looking to capitalise on the opportunity for margin expansion via vertical integration into downstream conversion, largely because of the price strength in lithium chemical products and the view that customers will want an ex-China supply chain.  Given Australia is long spodumene (and China is short) it makes sense to develop optionality around spodumene concentrate offtake and create a lever around which the "seaborne" lithium products markets can be kept tight.

Like extraction assets, processing assets will have different capital requirements and cost positions depending on their location, scale, and access to raw materials.  The cost curve for integrated processing assets remains in its infancy, given many projects that make up industry cost estimates are either still under construction or ramping up.

Figure 5 - IGO Corporate Presentation, 9 December 2020 

Supply

Ultimately to meet demand raw material supply will have to come from both hard-rock and brine assets.  Battery chemistry will vary depending on the availability of supply as well as the manufacturers preference, meaning supply of both lithium hydroxide and carbonate will be necessary. The main takeaway here is simply that the lithium battery supply chain is complex, and that given this it should be expected that theoretical supply will undoubtedly differ from realised supply.  Below is a consensus estimate of the future supply-demand balance for lithium (as measured in Lithium Carbonate Equivalent tonnes) out to 2030.  While obviously these estimates are rubbery, it gives a feel for the extent to which an imbalance may eventuate.

Figure 6- Airlie Funds Management

Price

Of course, the reason cost curves, and supply-demand paradigms are poured over by investors is to take a view on future prices.  Often with resources stocks, get the commodity price right and you'll give yourself a fighting chance investing in the right companies.  Over the past twelve months, prices for lithium products have exploded.

Figure 7 - BofA Global Research, January 2022

Figure 8 - Fastmarkets 2021 

But how should we interpret spot prices given the structure of the lithium industry? Are "spot" prices an accurate reflection of what producers are receiving? 

At present, spot transactions only account for a small portion of supply of lithium raw materials and chemical product producers, with most volumes traded in fixed price or index-linked contracts (for set periods).  For spodumene spot pricing, many market participants have begun relying on Pilbara Minerals (PLS) BMX Platform results, in which the company auctions off small parcels of spodumene to prospective customers.  If these "spot" prices reflect what is being paid for the marginal tonne of product, then they still give great insight into the market balance. 

Without wanting to oversimplify things or draw flawed comparisons, we can look at the development of the iron ore price (and value-in-use price variation) as a guide to how the lithium pricing system could mature.  The iron ore market moved to a spot-linked pricing system in 2008, despite considerable variation in product quality and specific end-use.  The result has ultimately been a significant increase in market transparency which we believe should be expected for lithium over the next decade.

Figure 9 - Fastmarkets 2021

With greater transparency over market pricing should come greater ability for market participants to allocate capital, and ideally create a smoother transition to electric vehicle use.  Yet, as investors we still must take a view on future prices even as pricing systems develop.  For mature commodities, long-term price forecasts typically reflect a marginal cost of production, where prices are set by cash operating cost levels that ultimately mean those at the top of the cost curve are not profitable, so as not to induce oversupply.  The general rule of thumb most people use here is equivalent to ~90% of the cost curve.  An obvious example where this logic is applied is to a mature commodity is again iron ore, where long run prices are usually US$60-80/t, with 90% of the cost curve effective profitable at ~US$70/t. 

Using this approach for say spodumene, would yield a long run price of ~US$450/t, versus spot of >US$3,000t.  Given the lithium market is not "mature" in the sense that pricing is underdeveloped, and the future supply-demand equation remains so unbalance, a marginal cost of production method for forecasting future price is perhaps unreasonable.  Instead, to address the future supply-demand imbalance predicted, new production needs to be incentivised - i.e., Long-term pricing must be bid-up to encourage investment in new supply, and so it's not out of the realm of possible that current spot prices can hold for longer than people expect, or for long run prices to settle above the current cost curve (especially given this cost curve will have to change over the next decade). 

All in all, without a crystal ball and given the plethora of unknowns, we remain open to the possibility that spot prices can hold or go higher despite their impressive run.  Even modest changes to the supply-demand equation can see hefty price responses, and it would be foolish to assume to future will not be volatile in both directions.

How are we navigating the lithium sector at Airlie?

Given the industry dynamics discussed above, we believe it is prudent to have some form of lithium exposure in our portfolio.  The uncertainty that features in all aspects of the lithium paradigm means each opportunity warrants a degree of conservatism, and valuation is still important, despite rubbery supply, demand, and price forecasts.  Undoubtedly, we will see an endless stream of new explorers-cum-developers front the market over the next decade, some of which will be fantastic investment opportunities and some of which will be looking to take advantage of investor optimism. 

In our previous Stock Story, we highlighted Mineral Resources (MIN) as our preferred lithium exposure.  Mineral Resources represents a compelling investment both in isolation, and when considering its relative valuation versus other ASX-listed producers.  For MIN, earnings growth will be driven by the organic expansion of spodumene and iron ore production volumes, as well as the development of a lithium hydroxide conversion plant via a JV with global producer Albemarle.  Supporting this growth is MIN's robust mining services business and exceptional management, and it remains a key holding in our portfolios.  

Written By Joe Wright, Analyst - Airlie Funds Management

Funds operated by this manager:

Webinar recording: Paragon Australian Long Short Fund

Some stuff you don't want to think about - but should

Delft Partners

January 2022

Or Risk is the independent variable and Return, the dependent one.

Ever been on a car journey as a passenger where the driver was going too fast; being reckless; and essentially taking chances? Over and under-taking, switching lanes without looking, going too fast and braking too hard. We all have. It's not so pleasant especially as you get older and realise how much is at risk from a needless mistake: your health, your life even; the realisation that family members depend on you staying healthy and not becoming a burden; your enjoyment from their growing old.

So, we try to avoid such a scenario. It's relatively easy since reckless driving is usually associated with inexperienced younger members of society in fast cars. So, we can avoid such journeys as being readily identifiable in advance.

And yet we often find that a similar journey, multi-generational wealth investing, has older wealthy members indifferent to risk and unaware of the loss of wealth that occurs when insufficient attention is paid to the risk path taken by their managers.

Investing Is not so different from a car journey? It's a path to an end point with risk to be managed; speed, comfort, time horizon all matter for investing as much as they do for being driven. Better to get to the end point safely and aware of risks on the way, and how to drive defensively at such times, than to get there at speed dangerously?

And yet here we are, perhaps caused by years of unorthodox monetary policy, where many managers and clients perceive risk as irrelevant and an irritant? All focus is upon return - those 'sexy' growth stocks, that 10 bagger PE fund, that property that will inevitably benefit from a lick of paint or a change of use permission, and of course the free put option available to all investors that has been a feature of monetary policy in essentially the last 25 years.

So were going to bore you with a brief article on the independent variable known as risk. Return, the glamourous bit, is the dependent variable. No risk, no return. So, if you can't define, measure, and manage risk then your returns are a function of luck. Not a lot of people know that.

We spend a lot of time obsessing about risk especially in multi-asset portfolios, whose Alphas and Betas are both correlated at different extents at different times, and where they offer differing opportunities for active and passive approaches at different times.

The cost of variance or variance drain

The variance of returns over time directly impacts our ability to grow wealth. $100 invested at 5% interest for 30 years grows to become $432. Of that, $100 is of course your initial capital, $150 is the total amount of simple interest paid, and $182 is the effect of compounding - interest on interest. (Let's assume away the taxman?!)

Once uncertainty in returns (volatility) is introduced into the mathematics, there is an immediate noticeable drain. If that same investment still yields 5% but now with a volatility of 17% (like a typical equity index ) the cumulative growth rate - the rate your wealth actually grows over time including rises and falls - is now just 3.56%. After 30 years you will have instead of $432 just $285, and you will have lost almost $150 to volatility. Risk is expensive!¹ A lower-return investment with a lower risk may often outperform a high-return high-risk alternative.

Multi-asset portfolios of course mitigate the high volatility from equity and many investors choose hard assets to go into the mix since their price is 'not volatile'. However, bonds, especially high yield bonds, can behave like equities at certain times ² and an absence of price volatility shouldn't be equated with an absence of risk ³. For example, if you own a physical property whose revenues are derived from retail tenants, and your equity portfolio is heavily weighted to retail and consumer companies, guess what happens when a general consumer recession hits. That 'lack' of volatility in the property will suddenly not be so lacking as your cash flows from rental payments dry up and the next valuation of the property will be a jump down in value, and a reduction in re-investable cash flows, just when you most needed to be immune from volatility as a drain on wealth.

The infrequent tails that happen too frequently

We tend to think of "drawdowns", crashes, corrections, shocks or other sudden unusual "non-Normal" events as a completely different animal from typical market volatility. When we consider investments, we should consider not just volatility or tracking error but also the incidence of these drawdowns. Simple mathematics to transform mean and volatility without drawdowns to equivalent numbers that correctly include the effect of such infrequent drawdowns has been known for almost 100 years ⁴. If we make this adjustment to historical return series, we observe two troubling effects. First, the historical "alpha" that is often reported in academic papers and relied on by active managers from factors like Momentum, Size, Value, or Beta, once adjustments for drawdowns have been made, either radically shrinks, or disappears, or reverses sign and becomes a persistent drag. This phenomenon is described at length in one of Northfield's recently published journal articles ⁵. Secondly, adjusted volatility numbers are usually much bigger than the unadjusted numbers - meaning that these investments have both a lower return and a higher risk than would at first appear. The effect of that higher risk is once again, as above, to reduce the effectiveness of compounding. Investments that experience periodic shocks have both lower return than would appear at first glance, and higher risk. Drawdowns are also expensive!

The evasive Alpha

Another closely related aspect is uncertainty in active management. Unfortunately, Alpha is hard to find, easy to lose, and always changing.

Because Alpha is uncertain, we should include that uncertainty in our calculations of risk - in this case it is "the risk of being wrong" or strategy risk ⁶. It may come as some surprise that usual risk measures completely ignore this and assume that the managers' return is known with certainty. We wish it was. With active management we deliberately move away from the benchmark return and believe that we have a new mean return (our alpha) that is bigger than the benchmark. Tracking error, which originated in passive management, or value at risk, or any other measure that is based on volatility measures variability or uncertainty around the mean (alpha) and says nothing about variability or uncertainty of the mean (alpha) itself. As with all the previous examples - you might be spotting a pattern here - correct inclusion of strategy risk makes the total actual risk of the investment larger, which once again negatively impacts compounding. Active management where big bets are taken is risky and being wrong is very expensive indeed!

There is no such thing as linear in the world of risk

There is one additional implication of compounding that is utterly ignored - if we consider the compound return, as we should, given its importance to us, then return no longer increases linearly with risk as we are told by the textbooks ⁷. It is now a convex function, and as risk increases past the peak, our return begins to decline with increasing risk.

This has serious implications for asset allocation. In every market there will be a "peak beta", beyond which as explained returns decline with increasing risk, so we must stay to the left of that peak. This tends naturally to send us looking for lower beta stocks, or into multi-asset class portfolios ⁸.

So, what to do?

Experienced judgement helps. By experienced we mean professionals whose careers extend back to when corporate failure was possible, when central banks tightened monetary policy as much as they engaged in QE and granted a free put option, and when dividends and balance sheets mattered.

Models also help to provide a framework for identifying, measuring, and managing the Betas and the Alphas and their correlations and payoffs. Many fund managers do not use risk models and are ignorant of the absence of diversification of the risk sources in their portfolios. Many don't want to know and prefer a 'pedal to the metal' approach.

We acknowledge that all models are 'wrong' but believe that some are useful. The useful ones such as the Northfield model we use (northinfo) allow liquid and illiquid asset forecasts to be combined; have data on bonds as well as equities, allow scenario and stress testing, and incorporate the multi-generational time horizon and planning for periods of investment and consumption that many plan sponsors will have to manage.

We conclude that successful wealth management requires carefully nurtured compounding. This compounding can only occur when three forms of risk are carefully managed the exposure to volatility from risk factors and asset-specific sources, the risk of drawdowns, and strategy risk - the risk of being wrong.

Portfolios are constructed by focusing risk exposures only in areas the manager has skill.

Funds operated by this manager:

Delft Partners Global High Conviction Strategy, Delft Partners Asia Small Companies Strategy, Delft Partners Global Infrastructure Strategy

• ¹ E.g. www.morningstar.com/indexes/xasx/xko/risk
• ² It's not hard to understand why - equity and bonds are issued by the same underlying entity. If the entity is relatively well financed and stable interest rate risks will dominate, however if the entity's future is less certain then credit risk concerns will dominate. Obviously the same factors that drive equity risk also drive credit risk - it is the same underlying business of course. See Merton's paper https://www.jstor.org/stable/2978814
• ³ Northfield's award-winning real estate model looks beyond appraisals to avoid this problem. https://www.soa.org/globalassets/assets/files/resources/essays-monographs/2014-erm-symposium/mono-2014-erm-belev.pdf
• ⁴ See for example "Moments and Cumulants in the Specification of Distributions", by E. A. Cornish and R.A. Fisher (1937), or Edgeworth expansions: http://personal.psu.edu/drh20/asymp/fall2004/lectures/edgeworth.pdf
• or Maillard, Didier, A User's Guide to the Cornish Fisher Expansion (May 1, 2018). Available at SSRN: https://ssrn.com/abstract=1997178 or http://dx.doi.org/10.2139/ssrn.1997178
• ⁵ Kantos, DiBartolomeo "How the pandemic taught us to turn smart beta into real alpha" Journal of Asset Management (2020) 21:581-590 https://doi.org/10.1057/s41260-020-00195-w
• ⁶ See: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=569281
• ⁷https://www.investopedia.com/terms/s/sml.asp
• ⁸https://www.northinfo.com/documents/576.pdf

Airlie Market Outlook, January 2022

Airlie Funds Management

17 January 2022

Matt Williams, Portfolio Manager, offers his views on the year ahead, the challenges he sees for Australian companies and discusses investments in the portfolio. 

Funds operated by this manager:

Peer Groups

FundMonitors.com

Why you can be cautious on markets and 100% invested

Aoris Investment Management

14 January 2022

Funds operated by this manager:

Aoris International Fund

Sitting On The Mountain, Watching The Tigers Fight 

Arminius Capital 

11 January 2022

Distinctively different and driven to perform Unlocking

the opportunity of 'special situations'

Colins St Asset Management

January 2022

Michael Goldberg (Managing Director / Portfolio Manager) and Rob Hay (Head of Distribution & Investor Relations) from Collins St Asset Management share insights into how 'special situation' investment opportunities have been implemented within the Collins St Value Fund.  Recorded 27 October 2021.

Mineral Resources

Emma Fisher, Airlie Funds Management

12 November 2021

Funds operated by this manager:

10k Words - January Edition

Equitable Investors

January 2022

Funds operated by this manager:

Equitable Investors Dragonfly Fund