Between end-2016 and end Q119, the (three-month) tin price traded in a relatively narrow range of US$20,131/t (±10%). From there, it declined materially (15%), to US$17,175/t by the end of December 2019, following a reduction in tin demand as a result of the effect of the US/China and Japan/South Korea trade wars on the global electronics industry. It then fell further (23%) to reach US$13,250/t in early 2020 due to the disruption to the world economy on account of the coronavirus outbreak and, in particular, the initial decline in demand from the world’s largest consumer, China.
From its nadir of US$13,250/t however, the price has to all intents and purposes, doubled as, first, producing countries such as Indonesia and Malaysia began to impose their own coronavirus restrictions, thereby constraining supply, and then demand began to recover in China as it started the process of relaxing its restrictions. As a consequence, the market is now in its tightest squeeze in over three decades. Refined tin exports so far in 2021 have recovered to pre-coronavirus levels at the same time as the world’s top exporter, Indonesia, is struggling to ship product as anti-coronavirus restrictions at its mining and smelting operations in Bangka Belitung (Indonesia’s largest tin producing province; NB it was formerly spelt Billiton and is the origin of the mining company of the same name) curb supply, while demand for solder and electronic products has risen, rather than fallen, as a consequence of the pandemic. As well as disrupting mining operations, seasonal bad weather has also disrupted exports of tin product from Indonesia, with January shipments from Indonesia falling 35% compared with December. More generally, like other tin-producing regions of the world, Indonesia is suffering from a depletion of reserves and, at the same time, is enforcing a crackdown on small scale and artisanal mining.
Exhibit 4: Cash LME tin price (99.85%), US$/tonne, January 2020 to present
|
|
|
Note that we recently published a more detailed description of the longer-term dynamics driving the tin market (Edison explains: Tin), including the effects of the electrification of the world economy and transport systems and high-tech applications such as robotics on likely global demand. To the extent that these result in a higher future real price of tin, the following analyses may prove conservative (our Alphamin valuation sensitivity to the long-term price of tin is provided in the ‘Sensitivities’ section of this note).
Edison has adopted two methods in determining its long-term tin price. The first is its long-term real price. The second is the correlation of the real tin price with the real oil price.
Relative to its crustal abundance, the price of tin is something of a statistical anomaly.
Conducted at the same time that our report, Gold stars and black holes, was published in January 2019, a regression analysis between the logarithm of the price of a metal or mineral (in US dollars per tonne) and the logarithm of its crustal abundance demonstrated a strikingly close correlation.
Exhibit 5: Graph of log (price in US$/t) vs log (crustal abundance in ppm), selected metals and minerals
|
|
Source: Edison Investment Research
|
In this case, the Pearson Product-Moment (correlation) coefficient of the regression is -0.97 which is statistically significant at the 5% level (that is to say, there is less than a 5% chance that the relationship occurred by chance). The position of tin (currently) on the graph is shown by the arrow. In this particular case, however, it can be seen that the current price of tin is low relative to its crustal abundance of 2.2 parts per million (ppm). In fact, on the basis of the strict correlation between the two, a crustal abundance of 2.2ppm should, all other things being equal, imply a price of US$125,265/t with a likely lower limit (derived from the error of estimation) of US$39,811/t. The fact that tin is trading at a price of US$26,215/t at the time of writing and that its price has only ever peaked at US$34,700/t in nominal terms and US$63,429/t in real terms therefore indicates that it is cheap in relation to its crustal abundance. At the same time, the following chart demonstrates that tin is also over-exploited relative to its crustal abundance:
Exhibit 6: Crustal abundance (ppm) vs annual production (tonnes), selected metals and minerals
|
|
Source: Edison Investment Research.
|
The explanation for these two apparent anomalies is likely to lie in the fact that tin – on account of its very high specific gravity – is relatively easy to concentrate and extract from its ore from a metallurgical perspective. Nevertheless, it has potential consequences for the long-term price of tin in the event that its continued exploitation at current rates depletes existing reserves and resources. Within this context, it is especially interesting to note that (uncommonly among natural resources) the countries that account for the highest proportion of global production are not the same countries that account for the highest proportion of resources (or reserves).
Exhibit 7: Percentage proportion of global tin-in-concentrate production (%), selected countries
|
Exhibit 8: Percentage proportion of global tin resources (%), selected countries
|
|
|
|
|
Exhibit 7: Percentage proportion of global tin-in-concentrate production (%), selected countries
|
|
|
Exhibit 8: Percentage proportion of global tin resources (%), selected countries
|
|
|
Notable discrepancies between production and resources can be observed for China (27% of global production, but <8% of global resources), Indonesia (26% of production, but <8% of resources) and Myanmar/Burma (17% of production, but <8% of resources) and, at the other end of the spectrum, Australia (15% of resources, but only 2% of production) and Russia (29% of resources, but <1% of production). In broad terms, while global production is concentrated in mid- to low-income countries (eg China, Indonesia, Myanmar/Burma), global resources are concentrated in mid- to high-income countries (eg Australia, Russia). This too may become significant if there is a shift in production from the former to the latter and the effect that this could have on the cost curve and the marginal cost of production. Note that a potential example of this shift may already be apparent in Myanmar/Burma, where concentrate supplies to China (which produces c 50% of the world’s tin) are reported to have fallen by c 30% in recent months as immediately accessible ore from unregulated mines has been exhausted.
Notwithstanding its over-exploitation relative to its crustal abundance however, the absolute level of consumption is still very low at only 46.2g tin per person per year. As such, even a small increase in consumption in grams per head per year in the future will translate into a requirement for materially more tin production globally in percentage terms.
Since 1945, the average real price of tin has been US$23,425/t, with a peak in 1980 of US$63,429/t and a trough in 2002 of US$9,500/t. The standard deviation of this population of prices is relatively high, at US$11,578/t. However, there is a valid argument to say that prices were squeezed upwards artificially during the 1970s by the actions of the International Tin Council (ITC) and that the subsequent collapse in price between 1980 and 1986 was equally artificial. More details of the history of this episode may be found in the Appendix at the back of this report, on page 21. Nevertheless, the shock of the ITC collapse and its aftermath may be clearly seen in the following graph of the real tin price since 1945:
Exhibit 9: Real tin price, 1945-present (US$/tonne)
|
|
Source: USGS, Independent Oil & Gas Association, US Bureau of Labor Statistics
|
In addition to the impact of the ITC’s failure on the price of tin, a number of other features of the graph are noteworthy:
■
Excluding the period 1973–2002, the real price of tin has rarely been above US$30,000/t and only very rarely below US$16,500/t; in the period 1945–1972, the average real price of tin was US$23,354/t; in the period 2003–2020, the average real price of tin was US$23,081/t (ie less than 2% difference between the two periods); over the whole period from 1945–2020, the average real price of tin was US$23,425/t.
■
Although the standard deviation of the population of real price numbers is high, at US$11,578/t, it contracts to a much more reasonable US$5,585/t if the anomalous 1973–2002 period is excluded.
■
The third noteworthy feature of the graph is the correlation between the real tin price and the real oil price (see below).
Real tin price: Real oil price correlation
The prices of many commodities have a close correlation with the price of oil. The qualitative justification for using the correlation as a means of calculating a long-term metal price is that energy costs make up a large proportion of a typical mining company’s total costs and so determines the absolute level of the cost curve. Should metal prices deviate materially from their oil price correlation for material periods of time therefore, it would result in mining companies being exposed to either excessive profits or losses which, in the case of the former, should result in an increase in investment, supply and/or substitution and, in the case of the latter, would be unsustainable. Within this context, it should be noted that Alphamin itself is an exception, given that its gravity and water-based concentration process is extremely energy efficient and not at all intensive. More generally however, the methodology may be justified on the basis of the energy intensity of the smelting and refining processes for tin, which tend to take the form of either electrolytic purification or, more typically, smelting in a traditional furnace.
Exhibit 10: Real tin price (US$/t, 2021 money) versus real oil price (US$/bbl, 2021 money), 1945–2020
|
|
Source: USGS, Edison Investment Research, US Bureau of Labor Statistics, Independent Oil & Gas Association
|
Between 1945 and 2020, the Pearson Product-Moment (correlation) coefficient between the real tin price and the real oil has been 0.55 (in a scale between -1 and +1). While that is statistically significant at the 5% level for the number of data points employed in the analysis, readers should nevertheless be aware that this has not always been the case. The graphs below show the correlation between the real oil price and the real tin price over time:
Exhibit 11: Real oil versus real tin Pearson Product-Moment (correlation) coefficient, 1945–2020
|
|
Source: Edison Investment Research. Underlying data: USGS, US Bureau of Labor Statistics, Independent Oil & Gas Association.
|
While the correlation between the real oil price and the real tin price was strong in the immediate aftermath of the Second World War (the green line in the above chart), it is notable that the correlation waned to nothing (ie uncorrelated) by the late 1960s. However, it was re-established strongly once again by 1974 and has remained so ever since with the brief exception of the period 2015–2017 (the grey line in the above chart).
Within this context, on the basis of a long-term real oil price of US$60/bbl (that currently being used by Edison’s oil & gas team), the long-term real tin price should be US$27,555/t. We believe this is an eminently defensible long-term tin price to use in our valuation of Alphamin. For reasons of conservatism, however, we have elected to use the (lower) long-term average, real price of tin of US$23,425/t. Note that variations from this level are considered in the ‘Sensitivities’ section.