BHP’s greatly overvalued because copper’s been hugely overhyped

Overview

BHP Group (hereafter “BHP”) is the world’s largest (by market capitalisation) mining company; by the same metric, it’s also Australia’s biggest corporation. It’s hardly certain, but it’s probably not controversial: its outlook is favourable. Reflecting its prospects, over the year to 30 June its shares surged more than 60%; indeed, they scaled an all-time high above $65. Three catalysts purportedly justify its upsurge.

#1: Copper’s “Super-cycle”

Its price soared more than 40% over the previous financial year, and to an all-time high exceeding $US13,000 per tonne. “Experts” attribute this surge to several strong tailwinds – such as the $5-7 trillion of capital expenditure that AI data centres and associated infrastructure are reportedly planning globally to 2030 (see, for example, McKinsey & Company, “The $7 trillion race for AI data center infrastructure,” 26 March). Executives at Nvidia are even more bullish: they expect that AI and associated spending will accelerate to $3-4 trillion annually by the end of the decade.

AI and data centre spending, as well as “climate action,” copper’s advocates enthuse, will necessitate a vast expansion of the world’s power grids, boost its demand for electricity and thus hasten its electrification. Furthermore, this enormous wave of capex will rank among the biggest in history. In the U.S., as a percentage of GDP it’ll rival the construction of canals and railways in the 19^th century, electrification in the 1920s and the development of the internet in the 1990s. Globally, it’ll thus require a colossal quantity of copper. Consequently, “experts” enthuse, it’s triggered a structural, demand-driven “super-cycle.”

#2: BHP’s Pivot towards “Future-Facing” Commodities

Copper is “future-facing” because it underpins the “global energy transition,” AI technology, and digital and other infrastructure. BHP is the world’s biggest producer of copper, and its latest half-year results, released in February, showed that for the first time the red metal contributed more than one-half of its EBITDA. If anybody benefits from copper’s alleged “super-cycle,” the crowd crows, BHP does.

#3: Copper’s “Constrained Supply”

The consensus is adamant: the world’s supply faces severe restrictions. Major financial institutions have forecast a shortfall of supply relative to demand up to 10 million metric tons (Mt) per year by 2040. Copper’s benefits to BHP, the consensus gushes, will therefore persist for decades to come.

In this article I dispassionately assess – and, as a result, strongly doubt – these claims.

Firstly, I demonstrate that copper’s “super-cycle” is an illusion. None of the four series of price data which I’ve analysed (I detail results from three) has ever – not even once! – experienced such a cycle. Over short-term, medium-term, long-term intervals since the late-1950s, copper’s CPI-adjusted price has regularly spiked upwards but nonetheless reliably reverted towards the corresponding rate of consumer price inflation, i.e., has remained roughly constant in “real” terms. Adjusted for CPI, copper is cheaper today than it was during the First World War; and since the first half of the 19^th century, the CPI-adjusted prices of most raw materials, including metals such as copper, have been essentially trendless – or have trended slightly downwards.

This result is hardly iconoclastic. Another analysis, which I’ll cite, concludes: “overall, (the prices of minerals including copper) have fluctuated up and down but around a reasonably consistent level. The time series are noisy, but the signal is that prices have not changed much over more than a century.”

Since the 1950s there’s been plenty of fluctuation, multiple cycles (regular but non-fixed fluctuations which persist longer than one year) but no “super-cycle” (rise of price which is statistically significantly above its long-term mean and which persists for more than a decade) has ever occurred. Nor do I uncover any evidence that one is now underway. Adjusted for CPI, today’s prices aren’t significantly higher than those prevailing more than 50 years ago.

“Experts” who assert that a “super-cycle” is occurring or imminent are ignoring consumer price inflation – and have been fooled by randomness.

Secondly, copper’s status as a “future-facing” metal is fictitious; the future which bulls confidently envisage simply isn’t emerging. Since 2000 fossil-fuelled electrification has occurred relatively rapidly in China and slowly in India; but electrification of any kind has advanced hardly at all in high- and low-income countries. Globally, it’s proceeding modestly; on a per capita basis, it’s progressing tepidly. Moreover, electricity’s share of the world’s total energy is rising sluggishly from a low base.

Most importantly, no “global transition from fossil fuels” is occurring.

Quite the contrary: in absolute terms, the consumption of hydrocarbons continues to rise – and in China, India and elsewhere it’s long been booming. As I’ll show, there’s no more evidence of a “global energy transition” than of Father Christmas and the Tooth Fairy. What I dub “strong form electrification” is merely a figment of climate zealots’ fevered imaginations.

Finally, for these and other reasons copper’s supply hasn’t been – and isn’t now – constrained. For more than 15 years, “experts” have been predicting that a shortage is (and thus price rises are) imminent. Yet over this interval CPI-adjusted prices haven’t risen significantly; hence there’s been no shortage.

Moreover, I doubt that $5-7 trillion of AI data centre capex will occur by 2030. The world’s biggest tech companies can no longer finance this capex from cashflows; hence they’re borrowing massively. Ever more studies are concluding that it’ll struggle to deliver significant returns, and that it could generate hefty losses. “The real-world costs of AI,” noted Forbes (“AI Costs More Than the People It Replaced,” 2 July), “are extraordinarily high while the payoffs are ambiguous.” Whatever’s unprofitable won’t continue.

Collectively, the results of my analysis indicate that bulls’ enthusiasm about copper in general and BHP in particular rest upon flimsy foundations – and imply that at some point disappointment – or worse – will ensue.

Bluntly, what bulls are spouting is at best highly questionable and at worst nonsensical. According to one astute observer, BHP is “just a liquid name for fund managers (most of them speculating) to ‘play’ the data centre theme.”

At its all-time high of ca. $65 in June, I estimate that its shares were more than 40% overvalued. A reasonable estimate of fair value is closer to $37.50, and for conservative (value) investors is closer to $30.

Caveats and Disclaimers

It’s important to emphasise: BHP is a world-class enterprise. Although bulls are grossly exaggerating its outlook, it’s nonetheless bright. It’s Leithner & Company’s kind of enterprise: not only does it lead its industry; it produces vital goods, has reliably generated profits and dividends, and possesses the managerial depth and fiscal heft to surmount cyclical and structural challenges.

We’ve held it in the past, and would happily – at the right price – do so again. BHP is a great company, but at $65 its shares make a poor investment.

The analysis which follows is far from a comprehensive. I’ll focus upon what’s presently besotting the bulls: copper, “climate action” and electrification, and AI and data centres. I’ll therefore ignore BHP’s cash-cow of the past quarter-century (iron ore) as well as its mooted third pillar of the next quarter-century and beyond (potash); I’ll also omit virtually all details of Leithner & Company’s valuation. Instead, I highlight fundamentals which nobody else has emphasised or even mentioned. In these key senses, this article resembles several others:

• Why Santos and Woodside are among our biggest holdings (18 May);
• Speculators in gold are playing with fire (3 November 2025);
• Why investors needn’t – and presently shouldn’t – own gold (7 April 2025);
• Why CSL will likely underperform term deposits (9 March 2025); and
• A contrarian assessment of Macquarie Group (25 September 2023).

A Deep-Dive into Copper

No other material conducts electricity as efficiently as copper. Aluminium is much lighter and cheaper; hence it’s preferred for high-voltage, cross-country power transmission wires and supporting infrastructure. Copper’s electrical resistance, however, is far lower. It thus loses considerably less energy and heat; hence copper wires can be much thinner but still carry the same electrical load. 

Copper isn’t just essential; it’s irreplaceable: without it, as well as metallurgical and thermal coal, iron ore, petroleum and silicon, the industrialisation of the past 150 years – and consequent enormous increases of living standards and longevity – couldn’t have occurred. During the 20^th century, (1) the creation of electricity infrastructure, (2) urbanisation and suburbanisation and (3) mass production of consumer electronics drove the world’s rising demand for copper.

In 1900, the production and consumption of electricity was a localised luxury; by 1950 in high-income countries like Australia and the U.S., power generation had become a major national industry and consumption a necessity.

During the first half of the 20^th century, copper wiring for generators, transformers, etc., formed the backbone of what became mammoth electrical grids. In Australia and the U.S., urbanisation (ca. 1900-1950) and suburbanisation (ca. 1950-1975) necessitated immense quantities of copper; “re-urbanisation” (that is, redevelopment of inner- and middle-city areas as populations have grown and urban planning paradigms evolved since ca. 2000) is now doing likewise.

In the Western world from the 1950s to the 1970s, as standards of living improved and war-torn regions (namely Europe and Japan) first rebuilt and then expanded their economies, industry and manufacturing prospered – and thereby drove decades of strongly rising demand for copper. And beginning in the 1990s, demand from China skyrocketed.

In high-income countries during the 20^th century, ownership of petrol-powered cars became practically universal. They required large quantities of copper for starters, alternators, wiring and radiators.

Last but hardly least, a digital and telecommunications revolution commenced during the 20^th century’s final quarter. The proliferation of telecoms cables, computing hardware, handheld devices, etc., greatly boosted consumer-level demand for copper.

It’s reasonable to regard the developments summarised in this section’s second paragraph as megatrends – that is, fundamental shifts of economic, societal and technological landscapes. They last decades, create huge new industries, reshape existing ones and permanently alter economies and behaviours. These megatrends occurred primarily but not exclusively in what today are high-income countries.

As we’ll see in the next section, it’s reasonable to assume that in the decades to come these old megatrends will lose steam – and that much-touted new ones, such as AI and climate-related electrification, won’t replace them.

Quantifying Megatrends Should – but Likely Won’t – Curb Bulls’ Excitement

Annual global demand for copper grew from approximately 0.5 Mt in 1900 to 15 Mt in 2000. That’s a CAGR of ca. 3.5% per year. In nominal terms (that is, unadjusted for consumer price inflation), America’s GDP rose from an estimated $21.2 billion in 1900 to $10.3 trillion in 2000; that was a CAGR of 6.4% per year. Adjusted for CPI, it grew 3.8% per year over the century.

It’s a key fact which copper’s boosters overlook or deny: even with strong tailwinds blown by multiple megatrends, during the 20^th century the world’s appetite for copper grew little more than half as fast as America’s nominal GDP.

BHP projects that total global demand for copper will increase roughly 70% by 2050. That’s a CAGR of 2.2% per year over the next 24 years. (Crucially, this is BHP’s estimate – not mine; I’ll return to it below.)

If it’s correct – that is, even if the AI revolution and “global energy transition” unfold as their boosters expect – the demand for copper will grow more slowly during the first half of 21^st century than it did throughout the 20^th. This crucial implication, too, escapes bulls’ attention.

Figure 1 plots average rates of actual and expected growth (expressed as CAGRs) of copper consumption since 1900. Over various intervals between 1900 and 2025, global demand rose, on average, by 3.2% per year. During the two intervals to 2050, according to BHP, it’ll grow at an average rate of just 2.6% per year.

Figure 1: Global Consumption of Copper (CAGRs), Actual and Estimated, Selected Intervals since 1900

Why is it reasonable to expect that the consumption will grow more modestly over the next quarter-century than it did throughout the 20^th century? There are two reasons (I’ll detail the second one in the next section).

Waning economic and population growth is the most significant.

From 1900 to 2000, according to the International Monetary Fund (hereafter “IMF”) and net of consumer price inflation, global GDP rose approximately 3.0% per year. Conversely, between 2025 and 2050 its “real” growth is projected to average 2.1%-3.0% per year – and in high-income economies like Australia, just 1.5%-1.8%.

Crucially, India’s economy hasn’t grown nearly as rapidly as China’s – and in the decades to 2050, the IMF expects, won’t grow nearly as fast as China’s did in the 1980s-2000s. Indeed, like China’s, India’s growth will slow (Figure 2).

Figure 2: Average Rates of CPI-Adjusted GDP Growth, Actual and Estimated, to 2050

Globally, the IMF anticipates that GDP’s growth will abate in response to huge and rapidly-rising public debt, the sluggish increase (and in many countries, shrinkage) of working-age populations, the expansion of labour-intensive sectors such as healthcare and aged care, and thus stagnating and perhaps declining productivity.

China’s prognosis is even worse. It’s commenced a demographic crisis, i.e., a rapidly shrinking and ageing population. From the 2030s it’ll grow more slowly than the world – and much more slowly than India. Mostly for this reason, China will eventually grow old rather than get rich. Even so, it’s highly unlikely that India’s per capital income will ever exceed China’s (see also “Can India become rich before its population grows old?” 5 August 2024).

World Population Prospects, published by the United Nations Population Division, produces the most respected and widely cited projections. It reckons that China’s population will sag from the current ca. 1.4 billion to 1.25-1.3 billion by 2050, and plummet to ca. 640 million (and perhaps as low as 460 million) by 2100. This will surely qualify as a “megatrend” – and may rank among the most profound of the 21^st century. Among other things, significantly threatens China’s global ambitions.

Even if climate-related electrification and AI data centre and associated capex unfold as its boosters confidently expect, growth in other and much more significant areas will stagnate and fall. The net result – lower global GDP growth than during most of the 20^th century – implies decelerating demand for copper.

Copper’s Allegedly Constrained Supply

Copper’s bulls ignore or deny another key fact: for more than 15 years, “experts” have been predicting that a “structural deficit” is imminent. Their chatter has formed three distinct phases:

1. Early Warnings (2009–2015): a host of organisations, ranging from investment banks to the International Copper Study Group, purported to identify a “structural deficit.” ICSG’s prompted modelling from the UN Environment Programme – which predicted “massive” increases of demand and “severe” shortfalls of supply through 2050.
2. “Global Energy Transition” (2016–2024): as the transition to “renewable” energy allegedly accelerated, organisations such as the International Energy Agency sounded the alarm. They prophesied “massive” deficits of supply relative to demand, and that “decarbonisation” is “the primary driver” which underpins it.
3. AI and Defence Era (2025–Present): analysts including Goldman Sachs, Morgan Stanley, JPMorgan and S&P Global claim to foresee “structurally significant” and “severe” shortfalls of supply vis-à-vis demand.

Despite numerous warnings over the past 15 years, the wolf hasn’t appeared: as we’ll see (Figures 3-5), the price of copper has hardly budged in CPI-adjusted terms. A relatively stable “real” price hardly implies a constrained supply!

Why do I expect that the demand for copper ore will grow more modestly in the next quarter-century than it did throughout the 20th century? Slowing GDP and population growth provided the first reason; the rising importance of recycling supplies the second.

Recycling copper is very easy and highly profitable. Unlike plastics or paper, repeated reprocessing doesn’t degrade it. Indeed, it can be repurposed endlessly without decreasing its original conductivity. Accordingly, a massive copper scrap and recycling industry exists: globally, this market is valued at ca. $70-$74 billion and by 2030 is projected to surpass $114 billion; that’s a CAGR of 12.4% per year.

Processing copper scrap is eco-friendlier and more cost-effective than extracting and refining ore. Recycling saves up to 85% of the energy required to mine and process ore; recycled also retains up to 90% of the value of original copper. That’s why recycled copper currently meets ca. 30-35% of global demand; it’s also why BHP, McKinsey and others expect that this percentage will rise to ca. 50% by 2050. Thanks to recycling, 80% of all the copper that’s ever been mined remains in use today. That percentage, too, will likely rise over the next quarter-century.

Recycling boosts overall supply without requiring additional mining. Consequently, it isn’t just a vital source of supply: it helps to offset the declining quality of ores and the long development timelines for new mines.

Recycling also tamps upward spikes of copper’s “spot” price. The cost of recycling averages just one-tenth of the cost of mining and refining ore; hence an efficient recycling loop keeps production costs – and therefore copper’s market price – lower and more stable than they’d otherwise be.

Recycled and scrap is “the world’s biggest and most responsive copper mine.” The higher “spot” or scrap prices rise, the more incentivised providers become to bring recycled metal to market. Their actions relatively quickly inject more supply and thereby moderate rising prices and spikes.

Wherever possible, miners first extract the easiest-to-reach, highest-grade and lowest-cost ores. As they deplete, they proceed to less accessible, lower-grade and higher-cost deposits. How do mining companies combat rising costs? By applying ever greater quantities of ever-cheaper extraction and refining technologies: historically, they’ve lowered costs per unit produced, and – as I’ll demonstrate below – anchored prices.

Given these and other reasons, it’s sensible to infer that over the next quarter-century changes of the price of copper – like changes of the price of iron ore, etc. – will, adjusted for CPI, continue to do what they’ve done over the past half-century and more: remain roughly constant.

Analysing Copper’s Price since the late-1950s

It’s another salient fact which “experts” rarely admit: their ability to forecast copper’s price, especially over horizons longer than a few months, has always been very poor. Studies by organisations such as the World Bank conclude that consensus estimates and regression models are unable to foresee copper’s price, and even whether it rises or falls: they often fail to beat a basic “coin toss” or a simple “no-change” assumption (see in particular Forecasting Industrial Commodity Prices: An Assessment, April 2024).

Figure 3: “Spot” Price of Copper, CPI-Adjusted $US per Metric Tonne, January 1992-March 2026

Using data compiled by the IMF, Figure 3 plots copper’s CPI-adjusted “spot” price since January 1992. The Producer Price Index (Copper and Copper Products) is a monthly indicator of the prices received by domestic American producers for copper-base scrap and alloy bars, rods and wire, and copper-alloy plates, sheets and strip. The U.S. Bureau of Labor Statistics has compiled it since January 1967. It’s an index which I’ve adjusted for CPI and rebased to January 1967 = 100 (Figure 4).

Figure 4: Producer Price Index, Copper and Copper Products (January 1967=100), CPI-Adjusted, January 1967-March 2026

Finally, PPI (Metals and Metal Products: Copper Base Scrap) is a monthly indicator which the BLS has compiled since January 1957. It measures prices received by domestic producers for copper and brass scrap. It’s also an index; I’ve adjusted it for CPI and rebased it to January 1957 = 100 (Figure 5).

Figure 5: Producer Price Index, Copper Base Scrap (January 1957=100), CPI-Adjusted, January 1957-March 2026

Five key sets of results emerge from Figure 3, Figure 4 and Figure 5. Firstly, since 1957 copper has never experienced a “super-cycle.”

According to capital.com (12 July 2022), “a commodity super-cycle means an extended period of boom and bust in the commodities markets, with prices falling significantly above or below their long-term trends. These movements may even outlast the business cycle and typically persist for well over a decade.”

None of these three series have ever risen two standard deviations or more above their long-term means for more than a handful of months – and never for more than one year, never mind a decade. Moreover, this result has applied since the First World War (for brevity I’ve omitted those results).

Secondly, the longer is the interval the weaker becomes the linear trend. The slope of the best-fitting straight line in Figure 3 (R² = 0.42) is significantly upward; in Figure 4 it’s insignificantly positive (R² = 0.05) and in Figure 5 it’s trendless (R² = 0.00).

Thirdly, the longer is the interval the harder it becomes to claim that an upward trend is currently underway. The best-fitting curvilinear model in Figure 3 (order 4, R² = 0.62) does indeed imply that a cyclical upturn commenced in ca. 2020; however, the best-fitting curvilinear model in Figure 4 (order 5, R² = 0.49) implies that the cycle is currently cresting – and the best-fitting curvilinear model in Figure 5 (order 4, R² = 0.52) implies that it’s turned downwards.

Fourthly, in CPI-adjusted terms today’s (March 2026) prices aren’t high.

In Figure 3, the current spot price ($12,529 in March 2026) is lower than in May 2006 ($12,980). In Figure 4, the index in March 2026 (133.1) isn’t significantly higher than its mean from 2007 to 2011 (average of 119.8) or in June 1974 (132.9). And in Figure 5 the index in March 2026 (119.2) is identical to its mean from 1965 to 1974 (119.4).

Finally, it’s true that average CPI-adjusted prices and index levels have been significantly higher since the early-2000s than they were from 1975 to 2005. Yet average prices haven’t been significantly higher since 2005 than they were before 1975. In this crucial respect, copper resembles iron ore (for details, see Forget Next Year’s Commodity Prices: Focus on 2075’s, 1 September 2025). If this is a structural shift, it occurred more than 20 years ago, and has created three eras:

1. 1957-1974 (Post-War Boom): prices were buoyant and regularly spiked.
2. 1975-2005 (Long Decline): throughout the 1980s and 1990s and into the early-2000s, prices trended downwards. Thanks to advancements of mining technology and large discoveries, in the late-1990s and early-2000s copper’s CPI-adjusted price plumbed historic lows.
3. 2006-2026 (alleged “Super-Cycle”): beginning around 2005, what boosters have dubbed “a massive structural boom” altered the copper market; in reality, it’s merely reverted prices to their mean before 1974.

So much for a “super-cycle”! None has occurred in the past, and (for reasons I’ll detail in the next section) I also strongly doubt that one is happening today – or that one will occur during the next quarter-century.

Inferring Copper’s Price to 2050

In Forget Next Year’s Commodity Prices: Focus on 2075’s (1 September 2025), I demonstrated that nobody can reliably predict the prices over the next year of commodities which trade in liquid public markets – yet cautious analysis based upon sensible assumptions can plausibly infer them decades into the future. 

In the short-term (12-month intervals), prices can fluctuate violently. Over the medium term (five years), however, volatility tamps considerably; and over long and very long terms (ten and more years), it reduces still further.

In particular, factors such as unexpectedly strong demand or sudden disruption of supply can cause commodities’ short-term prices to skyrocket. During slowdowns, recessions and financial crises, on the other hand, prices often plunge. Prices’ short-term variability is so wide that “experts” simply can’t know – and if they were brutally honest, would candidly admit that they don’t know – what the price of coking and thermal coal, copper, crude oil, iron ore, LNG, etc., will be over the next year.

Whether it’s knowingly or unwittingly, they’re simply guessing. “There are two kinds of forecasters,” John Kenneth Galbraith acidly reckoned: “those who don’t know, and those who don’t know they don’t know.”

Three key results emerged from my analysis:

1. in the short term, nobody can reliably predict commodities’ prices;
2. over short periods, prices often fluctuate wildly and occasionally boom and bust;
3. over longer periods of more than 10 years and up to 60 years, CPI-adjusted prices change little; hence there are no cycles, never mind “super-cycles.”

Table 1, which presents mean CAGRs and their measures of dispersion (standard deviations) over various intervals, applies this analysis to the three series plotted in the previous section. It confirms that nobody can reliably predict next year’s prices – but anybody can plausibly infer them decades from today.

Table 1: CPI-Adjusted CAGRs, Three Prices of Copper, January 1957-March 2026

It tells us that copper’s CPI-adjusted mean CAGRs haven’t, as a rule, differed significantly from zero. As a result, during the next quarter-century it’s reasonable to expect copper’s “real” price, like iron ore’s, etc., will do what it’s done since the late-1950s: remain roughly constant.

Consider as an example the scrap series plotted in Figure 5. It contains 819 rolling 12-month intervals; the mean CAGR over those intervals is 3.2% and its standard deviation is 24.5%. This series also contains 771 rolling five-year intervals. PPI scrap’s mean CAGR over those intervals is 0.85% and its standard deviation is 9.6%, and so on for the series’ longer intervals and the two other series.

Reading down the columns, three key points emerge: as intervals lengthen, (1) the CAGRs’ means shrink consistently and cumulatively drastically; (2) their standard deviations (SDs) also shrink consistently and drastically; and (3) the SDs are virtually always (exceptions appear in bold red font) more than twice the size of the means.

Why are these results important? Firstly, in 12-month periods regardless of series, a comparison of CAGR and SD tells us that copper’s CPI-adjusted price might plunge, remain relatively stable or soar. Crucially, however, given the observations’ normal distribution and again comparing mean to SD, the mean doesn’t differ significantly from zero.

In the short-term, copper’s price has fluctuated greatly but in “real” terms hasn’t risen significantly.

Secondly, over longer periods and regardless of series, it’s increasingly unlikely that CPI-adjusted prices plunge or soar. Again, however, comparing each mean to its SD, over intervals of five years or more the mean doesn’t differ significantly from zero.

In medium, long and very long terms, copper’s CAGRs fluctuate ever less – and don’t differ significantly from CPI.

What about the statistically significant results? Notice that two of the three are very long-term (20- and 30-year) CAGRs of “spot” prices; notice as well (Figure 2) that these CAGRs necessarily start and conclude on each side of the enormous step-change of prices which occurred in the early-2000s. Unless another upward step-change occurs between now and 2050, these two results are anomalies, i.e., reflect a “one-off” phenomenon.

The third anomaly, the 20-year mean CAGR for the “spot” price since 2005, contains only three observations (CAGRs which conclude January 2026, February 2026 and March 2026). Each of them benefits from the relatively sharp increase (by historical standards) of “spot” copper’s price during 2026. In historical terms it’s unremarkable (Figure 6). On that basis, this small handful of observations provides an inadequate basis for inference.

Figure 6: CPI-adjusted “Spot” Price of Copper, Rolling 12-Month Percentage Change, January 1993-March 2026

Strictly (statistically) speaking, the CAGRs in Table 1, including very long-term ones, are random and thus unpredictable. On the one hand, they derive from approximately normal distributions; however, because their standard deviations – and thus their 95% confidence intervals – also shrink drastically, the range of likely results reduces greatly.

In this practical sense and over very long terms, copper’s very long-term CAGRs become “predictable” (note the quotation marks).

What about Electrification? “Climate Action”?

Surely by now, if not well before, copper’s and BHP’s boosters will have objected: “Chris, you’re ignoring the biggest megatrend of all: the global energy transition! In particular, you’re disregarding climate action – and consequently the world’s accelerating electrification. The massive growth of AI and data centres, as well as climate action, will necessitate an enormous expansion of the world’s electricity grids, boost global demand for power and thus hasten the world’s electrification. Over the next quarter-century and beyond, these things will require vast quantities of copper. As the world’s leading producer, BHP is perfectly placed to benefit hugely.”

Three Definitions

What is “electrification”? Its proponents seldom ask; I’ll adopt and investigate the implications of three definitions. By “weak form electrification” I mean an increase of the quantity of electricity produced from whatever source (i.e., hydrocarbons, nuclear, or solar, water and wind), both in absolute and per capita terms, over time.

I define “semi-strong form electrification” as an increase over time of the quantity of electricity produced from whatever source, both in absolute and per capita terms, which is more rapid than the corresponding increase of the quantity of hydrocarbons (i.e., coal, natural gas and oil) produced and consumed.

Finally, I define “strong form electrification” as the replacement of machines fuelled by hydrocarbons with electric alternatives powered by intermittent (“renewable”) energy. It means replacing coal, oil and gas-fired power stations with solar and wind farms; replacing petrol-powered vehicles with EVs; and using reverse-cycle air conditioners and electric heat pumps rather than gas or oil boilers, switching to induction cooktops and electric hot water systems, etc.

Strong form electrification thereby causes the consumption of hydrocarbons to decrease in absolute terms.

According to its most zealous advocates, strong form electrification slashes CO₂ emissions and thereby mitigates or even halts “man-made climate change.” Moreover, they insist despite a Mount Everest of disconfirming evidence, because electric machines are so energy-efficient and solar and wind power is so much cheaper than “dirty” alternatives, strong form electrification reduces energy bills. Going “clean and green,” its zealots insist, doesn’t merely save the planet: it also rescues household budgets!

#1: Weak Form Electrification

The evidence of weak form electrification is mixed but mostly favourable: it’s occurring relatively quickly in China, more slowly in India (the world’s most populous nation) and slowest in high-income countries (which consume disproportionate quantities of power); as a result, globally it’s progressing modestly.

Figure 7 plots data from the Statistical Review of World Energy and compiled by Our Word in Data (I’ve used its most recent dataset; some of its series lack observations for 2025). Globally, the world’s total consumption of electricity rose from 15,279 terawatt-hours (TWh) in 2000 to 31,772 in 2025. That’s an increase of 16,493 and a CAGR of 3.1% per year. That’s slightly below the average rate of global GDP growth during these years (approximately 3.3% per year).

Figure 7: Consumption of Electricity, Terawatt-Hours, 2000-2025

Over the decade to come, AI data centres’ share of global consumption of power will likely double – but from a very low base.

Currently on a global basis, AI data centres consume ca. 95 TWh of power, and total global consumption is 31,772 TWh. Hence AI data centres’ consumption as a percentage of the world’s total is currently 448 ÷ 31,772 = 1.4%. Accepting the International Energy Agency’s projection that AI data centres’ consumption will rise to ca. 1,200 GWh by 2035 and my assumption that global consumption continues to increase 3.1% per year, we have 1,200 ÷ 42,700 = 2.8%. In the U.S. the percentage will rise more slowly but from a higher base: from ca. 6% today to ca. 8-9% in 2035.

Globally, consumption of power is rising at a constant rather than an accelerating pace. It’s also crucial: the global growth of power consumption is primarily attributable to growth in China.

In that country, usage rose from 1,347 TWh in 2000 to 10,564 in 2024. That’s an increase of 9,217 (56% of the world’s total increase over the past quarter-century) and a CAGR of 9.0% per year. In India, consumption grew more slowly and from a much lower base: from 573 TWh in 2000 to 2,083 in 2024. That’s a lift of 1,510 (9.2% of the world’s total increase) and a CAGR of 5.5% per year. Finally, in high-income countries consumption increased from 10,772 TWh in 2000 to 13,013 in 2024. That’s a rise of 2,241 (13.6% of the world’s total increase) and a CAGR of 0.8% per year.

Is weak form electrification occurring on a per capita basis? Globally, it’s advancing tepidly: consumption rose from 2,476 kilowatt-hours (kWh) per person in 2000 to 3,860 in 2025 (Figure 8). That’s an increase of 1,384 and a CAGR of 1.9% per year – and is lower than global GDP’s mean rate of annual growth (3.5%) over these years.

Figure 8: Consumption of Electricity, Kilowatt-Hours per Person, 2000-2025

Unsurprisingly, increasing per capita consumption globally is attributable primarily to rising per person consumption in China.

In that country, usage of electricity rose from 1,061 kWh per person in 2000 to 7,460 in 2024. That’s an increase of 6,399 and a CAGR of 8.5% per year. In India, per capita consumption grew more slowly: from 541 kWh in 2000 to 1,423 in 2024. That’s an increase of 882 and a CAGR of 4.1% per year. Finally, in high-income countries per capital consumption has increased from 8,591 kWh in 2000 to 9,163 in 2024. That’s an increase of 572 and a CAGR of 0.3% per year.

#2: Semi-Strong Form Electrification

Is electricity’s share of total energy rising? Yes – but slowly and from a very low base (Figure 9). Globally, its share – remember, I’m indifferent to its source, whether it’s hydrocarbons, nuclear, water, wind or solar – rose from 12.7% in 1985 to 19.8% in 2024. That’s a CAGR of 1.1% per year; at this rate, by 2050 its percentage share will increase to 19.8% × (1.011)²⁴ = 25.7%.

Globally, electricity’s share of total energy is rising tepidly because China’s share is rising strongly: it’s increased from 7% in 1985 to 24% in 2024. That’s a CAGR of 3.2% per year. In high-income countries, in contrast, electricity’s share has grown from 14% in 1985 to 19% in 2024. That’s a CAGR of 0.8% per year.

Figure 9: Electricity’s Share of Total Energy, 1985-2024

China’s high rate of electrification ISN’T the consequence of a global “energy transition.” It reflects the world’s “industrial transition,” i.e., the massive shift of manufacturing from elsewhere, particularly high-income countries, to China.

Heavy industry and manufacturing consume ca. two-thirds of China’s power. It’s become the world’s foundry and factory, and thereby requires immense quantities of electricity to manufacture steel, cement, electronics, solar panels, wind turbines, batteries and much else (including consumer goods). Households consume 15%, and services and commercial establishments most of the remainder (ca. 17%).

Is electricity’s share of the world’s energy rising? Yes – but slowly and from a low base. Today, the world remains what it was in 1965: overwhelmingly fossil-fuelled (Figure 10).

Figure 10: Hydrocarbons’ Share of Total Energy, 1965-2024

Globally, hydrocarbons’ (i.e., coal, natural gas, oil and associated products) share of the world’s total energy decreased from 94% in 1965 to 81% in 2024. That’s a CAGR of -0.2% per year. At that rate, in 2050 hydrocarbons’ share of global energy will be 81% × (0.998)²⁴ = 77%.

China has decarbonised more rapidly. In that country, hydrocarbons’ share of total energy decreased from 96% in 1965 to 80% in 2024. That’s a CAGR of -0.4% per year. At that rate, in 2050 hydrocarbons’ share of China’s energy will be 80% × (0.996)²⁴ = 73%. High-income countries have decarbonised more slowly than China: electricity’s share has sagged from 93% in 1985 to 79% in 2024. That’s a CAGR of -0.2% per year. At that rate, in 2050 hydrocarbons’ share of energy in these countries will be 79% × (0.998)²⁴ = 75%. Finally, India hasn’t decarbonised at all: in 1965, hydrocarbons comprised 91% of its total energy; in 2024, they comprised 90%.

“Climate action” is thus a risible hoax and a cruel joke. “Net Zero by 2050” isn’t remotely achievable: it’s merely a source of huge subsidies for carpetbaggers and grifters. As such, it’s becoming a means which virtue-signalling people comfortably ensconced in sinecures fleece low-income earners. Tragically for the latter, crazed ideologues like Chris Bowen and Matt Kean take it seriously.

#3: Strong Form Electrification

There’s zero evidence in favour of strong form electrification – and overwhelming evidence which disconfirms it. If it were occurring, the world’s consumption of hydrocarbons would be falling in absolute terms. But it’s not falling. Quite the contrary: it’s rising (Figure 11).

Figure 11: Global Consumption of Hydrocarbons (1965=100), 1965-2024

Globally, consumption has risen from a base of 100 in 1965 to 352 in 2024. That’s a total increase of 252% and a CAGR of 2.2% per year. It’s true that in high-income countries consumption has decreased from its all-time high of 230 in 2006 to 215 in 2024. That’s a CAGR of -0.4% per year over this interval. From 1965 to 2024, however, consumption has increased 115% and at a CAGR of 0.3% per year.

Figure 12: Global Consumption of Hydrocarbons (1965=100), 1965-2024

Slowly decreasing consumption in high-income countries is inconsequential in a global context. That’s because in China and India it’s skyrocketing (Figure 12). In China, consumption has risen from a base of 100 in 1965 to 2,645 in 2024. That’s a total increase of 2,545% and a CAGR of 3.3% per year. In India, it’s risen from 100 in 1965 to 1,797 in 2024. That’s a total increase of 1,697% and a CAGR of 5.0% per year.

Note that the relationship in Figure 11 is very strong, and that the best-fitting model is linear. In plain English, there’s simply no evidence that global demand for hydrocarbons is decelerating – never mind decreasing. Also observe that the best-fitting lines in Figure 12 are curvilinear. The consumption of hydrocarbons in China and India isn’t merely increasing: it’s accelerating.

Finally, note that China’s and India’s quickening thirst for hydrocarbons, expressed as CAGRs, exceeds (1) the growth of copper’s supply and demand since 1900; (2) the growth of global GDP throughout the 20^th century; and (3) the growth of electricity’s supply and demand since 2000.

The implication is as obvious as it is fundamental: copper is a “future-facing” metal ONLY in the sense that it’s a necessary condition of further fossil-fuelled economic growth and development. Facing a future which isn’t appearing – that is, which presupposes “climate action” – is a contradiction in terms.

What about the $5-7 Trillion of AI and Data Centre CapEx to 2030?

According to McKinsey & Company (“The $7 trillion race for AI data center infrastructure,” 26 March), the world’s AI data centres and associated infrastructure will absorb $5-7 trillion dollars of capital expenditure by 2030. An economist at Columbia University, Stijn van Nieuwerburgh, is much more bullish: on 26 June he asserted that $8 trillion – nearly five times the market value of New York City’s property market – will be spent on the AI build-out in the U.S. alone by 2032.

In contrast, I doubt that this tsunami of “investment” will eventuate. Five challenges – which bulls overlook or deny – underpin my scepticism.

Firstly, the massive rollout of AI data centres requires skills and resources. Yet it’s not evident that they exist in the required quantities; indeed, supply chain and labour shortages have developed, and over the next several years they’re more likely to worsen than to improve. Secondly, and as already noted, the centres will require large amounts of power. Yet bottlenecks of supply are forming: grids are increasingly unable to meet existing centres’ demands, and lead times for essential equipment to power new centres have increased.

The relentless rise of power prices in high-income nations – above and beyond those caused by “cheap” intermittent sources of power – is one result. Meanwhile, utilities’ plans to pervade the grid across the countryside face growing hostility from local communities in these countries.

Thirdly, it’s reasonable to assume that technology will advance, and thus that tomorrow’s AI models will require less physical infrastructure and consume less energy than today’s. If so, the eventual need for centralised, hyper-scale “mega-campuses” won’t be as great as boosters confidently insist. Fourthly, AI chips, server and associated hardware are becoming obsolescent ever more quickly. This hastening “refresh cycle” threatens data centres’ returns.

Finally, and perhaps most importantly, tech companies’ infrastructure spending is exceeding their cash flows; accordingly, the AI buildout increasingly relies upon heavy borrowing – and a rapidly-growing mountain of debt. “What worries me the most,” said one of the IMF’s directors, “is there’s leverage by the borrowers and there’s leverage by the investors. The leverage on both sides is very worrisome for financial stability” (see “The World’s Top Economists Are Sounding the Alarm on AI,” The Wall Street Journal, 1 July).

Scepticism is rising that this frenzy of debt-fuelled capex will ever be profitable – and fears are rising that it’ll generate losses. Another economist told WSJ: “if AI is not going to deliver on all the promises of productivity gains, then a lot of the investments that are made today will … turn out to be … problematic.” That’s putting it mildly!

Ever more studies (conducted by reputable entities such Gartner, McKinsey, MIT NANDA and RAND, etc.), conclude that a vast majority of AI initiatives are failing to deliver meaningful returns. (see also Never mind DeepSeek: here’s why the AI mania won’t last, 3 February 2025).

Paul Krugman famously observed: “unsustainable situations usually go on longer than most economists think possible. But they always end, and when they do, it’s often painful.”

Anticipating and Debunking One Last Criticism

At this point, if they haven’t already, BHP’s and copper’s bulls will play their last card: “you need to look at growth forecasts in terms of volumes rather than growth rates!” Robert Friedland, the founder of Ivanhoe Mines, has over recent years repeatedly expressed a seemingly-startling claim: over the next quarter-century, humanity must produce as much copper as it has over the past 10,000.

Approximately 700 Mt of copper has been mined from prehistoric times to 2024. Friedland contends that, in order to lift global GDP’s growth to 3.5% per year, over the next couple of decades the world will require an additional 700 Mt (see, for example, “Robert Friedland says global economy faces copper crisis,” Canadian Mining Journal, 23 October 2025).

The IMF and others doubt that the world’s GDP’s growth will average 3.5% per year to 2050. Nonetheless, I reckon that Friedland’s estimate is likely cautious.

Given (1) current global output (ca. 29.4 Mt in 2025) and BHP’s estimate that (2) the total quantity demanded will rise 70% (i.e., to ca. 51 Mt in 2050) and (3) the contribution of scrap and recycled copper to total demand will rise to 50%, by my reckoning that’s total quantity demanded of ca. 987 Mt (of which recycled and scrap will comprise ca. 425 Mt) over the next 24 years. Yet this enormous increase DOESN’T portend rising prices.

Friedland claims that demand for copper is skyrocketing; relative to supply, I doubt that it is or that it will (recall Figure 1). More fundamentally, he’s hardly the first to foresee a (1) major imbalance of a commodity’s long-term supply and demand, and thus (2) a major and permanent change of its price.

These predictions’ accuracy hasn’t merely been abysmal; it’s been laughable.

Three (In)Famous Imbalances of Supply and Demand which Never Occurred

In 1972, for example, in a report entitled The Limits to Growth, the Club of Rome (an international think tank founded in 1968, composed of allegedly “leading” scientists, economists, business leaders and former heads of state) took as its starting point the global reserves of key commodities. It then divided them by consumption in 1970, assumed that no new discoveries or technological advancements would occur – and concluded that if consumption continued to grow exponentially then reserves would be exhausted in roughly 20-50 years (e.g., by 1993 for gold, 1997 for zinc, 2009 for crude oil and 2022 for copper).

In 1972, global reserves of copper were estimated at 300-350 Mt; today, they’re ca. 980 Mt.

The report’s alternative scenarios assumed that expanding technology and price increases would multiply known reserves by factors of five or more. In these “more realistic” scenarios, “widespread systemic collapse” was pushed into the 21^st century by pollution, food shortages or “failure of the ecosystem” rather than the physical exhaustion of minerals’ supply.

At the time of its release, The Limits of Growth was widely – and rightly – mocked. Competent economists dismissed it as a “doomsday fantasy.” Alas, for exactly that reason it’s been influential; in particular, it’s provided a template for the hysterical climate fearmongering of the past few decades.

A second famous example occurred in 1980. An unhinged lepidopterist (butterfly scientist), Paul Ehrlich, who in the early 1970s careened out of his lane and predicted that “hundreds of millions of people” would starve to death, that “all important marine life” would be extinct by the end of the decade and thus that average human life expectancy in the U.S. would collapse to 42 years, bet an economist and rational optimist, Julian Simon, that the CPI-adjusted prices of five metals – chromium, copper, nickel, tin and tungsten – would, as a result of rapid growth of population and physical depletion, boom over the decade to 1990. Simon contended that innovation would overcome scarcity, and thus wagered that these prices would fall.

From 1980 to 1990, the prices of all five commodities fell. Moreover, global deaths from famine plummeted, the world’s population of whales continued to recover and human life expectancy in the U.S. rose from 73.7 to 75.4 years.

Crucially, Simon’s point generalises beyond these minerals and the 1980s. During most decades and regarding most mineral and agricultural commodities, he would have won similar bets. Long-term human innovation, in short, overcomes apparent physical scarcity (see “Who would have won the Simon-Ehrlich bet over different decades, and what do long-term prices tell us about resource scarcity?” Our World in Data, 6 January 2025).

Finally, intermittently since the mid-20^th century, a few geologists, energy analysts and others have repeatedly warned that “peak oil” is “on the horizon.” They’ve predicted, in other words, that global output will crest and then collapse. An American geoscientist, M. King Hubbert, originated claim in 1956. The “Hubbert Curve” predicted that U.S. oil production would peak between 1965 and 1970. Since then, others have extended the concept globally.

In 1970, the U.S. produced an average of 9.64 million barrels per day (MMbpd). On an annualised basis, it produced roughly 3.5 billion barrels of crude oil that year. In 1970, the world produced approximately 45.89 MMbpd and ca. 16.7 billion barrels for the entire year. In 2025, the U.S. produced an all-time record average of 13.6 MMbpd. Including other liquids, its output reached an even higher peak of approximately 20.5-21 MMbpd. Last year, the world produced an average of approximately 106.1 MMbpd.

Why These Predictions Fail

Forecasters – climate catastrophists are merely the latest – consistently underestimate human ingenuity and adaptability. Acting through markets, people don’t merely imagine possible future circumstances: they’re constantly adjusting to current conditions. As a result, they devise innovations and solutions, exploration increases and substitution – as we’ve seen in copper’s case, recycling – occurs.

Why do extremists’ predictions such as the Club of Rome’s and Ehrlich’s, etc., fail so abysmally? Three reasons:

1. Prices Produce (Dis)Incentives: when a commodity’s quantity supplied at a given price falls short of the quantity demanded at that price, its price rises. A sufficiently higher price triggers a feedback mechanism: harder-to-reach deposits become profitable to explore, consumers seek alternatives, and industries alter and improve manufacturing processes in order to use less per unit of output. Over time these actions counteract the initial increase of price.
2. Static versus Dynamic Forecasts: predictions, whether pessimistic or exuberant, typically assume that current technologies won’t advance, and that production and consumption remain static. They fail to anticipate new technologies – and how they reshape supply and demand.
3. Resource to Reserve Ratios: they’re very poor predictors of prices. They measure only currently-identified and economically-recoverable deposits under today’s conditions – and NOT the identifiable and recoverable deposits under tomorrow’s.

In Knowledge and Decisions (1980), Thomas Sowell wrote: “the cavemen had the same natural resources at their disposal as we have today, and the difference between their standard of living and ours is a difference between the knowledge they could bring to bear on those resources and the knowledge used today.”

It’s therefore ironic; indeed, it’s blatantly contradictory. On the one hand, boosters of AI data centres celebrate recent rapid advances of technology, and assume that they’ll continue. Yet copper’s bulls – prominent ones also seem to gush about AI data centres – discount or ignore the impact of advances of technology upon BHP’s and other miners’ ability, present and future, to extract greater quantities of copper at constant or lower CPI-adjusted prices.

They can’t have it both ways.

Conclusions and Implications

Some funds managers – seemingly because they’re utterly ahistorical and lack the ability to reason – are presently gushing about copper in general and BHP in particular. One “is highly optimistic about where the resources sector is heading. ‘The next five or 10 years look extremely exciting for the mining sector. There’s going to be very little new supply in commodities like copper over the next five years. I don’t think the outlook for the mining industry has ever looked better’” (see “BHP will remain ASX’s most valuable company over CBA, say experts,” The Australian Financial Review, 19 May).

Another funds manager “expects copper to closely mirror the iron ore boom of 25 years ago when prices shot up tenfold. ‘Copper price can go up five times in the next five to 10 years, if not more,’ he said. ‘With the copper price high, this talk of electrification and obviously the building out of data centres, it looks like demand will stay strong for a while.’ The first manager predicts ‘BHP’s share price could skyrocket past $100 in the next couple of years,’ up from $58.82 on Monday. ‘Even then, we’d still be happy to be long,’ he said.”

These funds managers are merely speculators who’ve been utterly fooled by randomness. At its present price, I’m content that Leithner & Company is short.

What, in light of the points I’ve addressed – not to mention the myriad factors affecting the profitability of its production of iron ore and potash – might conservative, long-term investors reasonably pay for BHP’s shares? To this point I’ve omitted all details of our valuation. Figure 13 provides a snippet: it plots BHP’s cyclically-adjusted price to earnings (CAPE) ratio since the GFC (if you’re unfamiliar with CAPE, High Valuations – Not Rising Bond Yields – Threaten American Stocks (29 June) provides a description and discussion).

Figure 13: CAPE Ratio, BHP Group, January 2009-June 2026

BHP’s CAPE has been strongly cyclical over short and medium terms, but trendless – and thus heavily mean-regressing – over the long term. It currently (June 2026) exceeds 25, and its mean since the GFC is 17.5. Stripped of many complications and details, this mean helps to provide a reasonable estimate of its shares’ value.

A regression of BHP’s CAPE to its mean implies (given assumptions which, for the sake of brevity, I’m omitting) a decrease of the price of its shares to ca. $37.50. It suggests, in other words, that at $65 these shares were more than 40% overvalued, and to return to fair value they must suddenly plunge or gradually sag more than 40%.

What does this mean? Certainly NOT that price of BHP’s shares will shortly plunge 40% or more.

In Stop kidding yourself: Nobody can “time the market” (30 June 2025), I demonstrated that any attempt to “time the market” (speculate) is virtually impossible. Short-term market returns – I used BHP’s 12-month returns as an example – are essentially random and thus unpredictable.

Hence nobody can plausibly foresee the price of BHP’s shares a year from now. They could be lower; if so, they’ll be less overvalued. They could also be higher – in which case they’ll be even more overpriced.

What, then, does BHP’s overvaluation imply? It means that at $65 the likelihood is that BHP’s shares will generate a poor total return over the medium (next five years) and long-term (next 10 years).

Investment, like bridge and chess, is ultimately a matter of assessing probabilities. For owners of BHP’s shares at $65, the odds are unfavourable; for purchasers at ca. $37.50, they’re reasonable; and towards $30, they’re attractive.

Buyers of BHP at $65 credulously accepted the bulls’ “narrative” and discounted, ignored or denied the data and analysis I’ve detailed in this article. They’ve accentuated a rosy future and attenuated the past and present. They’ve ignored the steak – and let its anticipated sizzle bamboozle them. They’ve succumbed to emotional appeals, abandoned scepticism and overlooked or denied fundamentals; that, to put it mildly, is hazardous.

In December 2021, Livewire “surveyed more than 4,000 investors and learned that 62% of them intend to invest in decarbonisation in 2022. That was double the level of interest of any other megatrend.” Its “Megatrends Series,” which it dubbed Decarbonisation 2022, unearthed “the funds, ETFs, and experts leading the charge to Net-Zero.” One of its contributors declared: “climate represents a significant long term investment opportunity.” That assessment was, and remains, absurd.

For conservative investors, decarbonisation wasn’t and isn’t a long-term opportunity: it’s yet another ephemeral mania – such airlines in the 1950s, semiconductors in the 1960s and the internet in the 1990s – which reliably evolved from boom to bust (for details, see Decarbonisation: A doubter’s guide for conservative investors (16 May 2022, and “Net zero” isn’t a Megatrend: It’s a Mega-trap, 8 April 2024).

From a conservative (that is, value) investor’s point of view, copper’s “super-cycle” and AI data centres are merely the latest in a long series of passing crazes.

The Bank for International Settlements reckons than an AI bubble and bust rank among the most alarming threats to global prosperity. In the Annual Report it released on 28 June, it warned: “disappointment in returns could trigger a sudden pullback in financing and turn the (AI) capex boom into a protracted investment bust, with potential knock-on effects on financial conditions.” Indeed, the “repricing of risk, whether triggered by higher interest rates or an AI bust, has the potential to (trigger a severe global credit crisis).”

My sceptical assessment doesn’t alter the most fundamental fact: BHP is a world-class company. To be clear: regardless of its shares’ price, BHP’s future is bright; at $65, however, the outlook of its shares’ returns is dim.

“There’s no commercial institution in Australia that has contributed (as much as BHP) to the nation,” declared Geoffrey Blainey, an historian of Australia and its mining industry, on the 130^th anniversary of its founding. “Three of the big banks are older, but you can’t say (any of them) has contributed as much” (“Blainey Hails BHP’s 130-Year Contribution,” The Australian, 14 August 2015).

Given its past achievements and future prospects, I expect that BHP will remain what it’s long been: (1) a leader of its industry; (2) a provider of essential goods; and (3) a reliable generator of profits and dividends which (4) possesses the financial strength and managerial depth to address various “headwinds” (cyclical and possibly structural) and pursue opportunities.

Leithner & Company has held BHP before, and at the right price would gladly do so again. Its strengths and opportunities are formidable. Yet it’s easy to overestimate them, and by grossly overhyping copper today’s bulls have succumbed to the temptation. As a result, they’ve greatly overvalued its shares.

Click LIKE so that Livewire knows that you want more of this type of content. Click FOLLOW for notification when my next wire appears.