copper market primer

European Metals & Mining: Copper for the Craftsman Cunning at His Trade

SEPTEMBER 2013

SEE DISCLOSURE APPENDIX OF THIS REPORT FOR IMPORTANT DISCLOSURES AND ANALYST CERTIFICATIONS

US$10,000/t copper...there’s no substitute for quality and this is in short supply

Is US$10,000/t copper a possibility? Yes, it is — this Blackbook examines the forces that will send the copper price to new highs, with 2016 seeing copper begin its ascent towards US$10,000/t

Chile played a unique role in providing the world with two decades of copper at suppressed prices; but it will be impossible to ever replicate the impact of Chile on copper; sources of supply not economic at today's price must soon be brought into play

In the face of increased geopolitical risk, falling grades and the declining success of exploration, significantly higher copper prices will be required to incentivize the necessary new mine capacity

The total amount of copper embedded in China's capital stock today is 29kg/capita — equivalent to just 30% of the Japanese levels and 22% of South Korean levels; as a later-cycle commodity, copper demand in industrializing China has a long way to go

For the exclusive use of JASON LAPORTE at PERRY CAPITAL on 25-Sep-2013

EUROPEAN METALS & MINING: COPPER FOR THE CRAFTSMAN CUNNING AT HIS TRADE 1

Portfolio Manager's Summary

Our analysis is complex, but the thesis is simple: We do not believe that the copper price has become detached from the economics of supply and demand. Rather, we believe that the rising copper price over the last decade has triggered five effects critical for maintaining the supply/demand balance. First, it allowed for the entry of 1Mtpa of high-cost sub-scale material from China. Second, it enabled the exploitation of lower-grade sections in existing mines, thus offsetting the effects of the deposits' geological degradation. Third, it incentivized investment in new projects, despite the declining exploration success and increased geopolitical risk. Fourth, it accelerated the use of scrap copper as a substitute for mined material. Finally, it has witnessed the replacement of copper in nonessential applications with aluminum and steel. This Blackbook is primarily supply side focused and examines the first three of these effects.

Our thesis can be summarized in five key points. First, copper is the most overutilized major commodity, and a higher price should be the result of an overexploited metal. Incentivizing an increase in the supply of a metal with little availability and an already high rate of exploitation requires prices to move upwards. Second, major new discoveries are declining. The rate of new large copper porphyry discoveries has ground to a halt. In contrast, demand has accelerated on the back of Chinese industrialization. Third, China's importance as a copper supplier is rising. In terms of growth, the last decade has seen Chile stall and China emerge as the fastest-growing source of supply — currently the world's second-largest supplier, despite a lack of high-quality geology. Fourth, the missing 1Mt of Chinese production is very high cost and will only get more expensive over time, thus further steepening the cost curve. Fifth, grade declines increase costs enormously. Although the declines in global head grade may be well known, the implications of this on future copper prices are not. Because head grades are too high today, they must fall tomorrow.

Of the movement in copper, 90% is explained by just three factors: grade, GDP and inventories. A regression — in the form of Price = + 1 x Grade +2 x GDP +3 x Inventory — explains nearly 90% of the variation in copper price over the last 35 years. By far, the most important element is grade. We can then use this relationship as the basis of a price forecast. Even the most conservative grade profile gives rise to prices well in excess of US$10,000/t. It is only a question of when this occurs, not if, in our view.

All the major miners have copper exposure but none more so than Glencore Xstrata, which stands to benefit the most, should our forecast hold true. In particular, we like the decisiveness of Glencore Xstrata in its commitment to the development of the "new world" copper assets in Peru and Central Africa. For us, the most significant new copper development belongs to Rio Tinto, with Oyu Tolgoi in Mongolia. We also note that Rio Tinto has a number of longer-dated growth options in La Granja and, especially, Resolution. In addition, we remain "believers" in the copper price story. In our view, the fundamentals for copper are conducive to the price staying genuinely "stronger for longer" relative to, say, iron ore.

Paul Gait [email protected] +44-207-170-0599Christian Cole [email protected] +44-207-170-5101Rusne Didziulyte [email protected] +44-207-170-0541

September 24, 2013


2 EUROPEAN METALS & MINING: COPPER FOR THE CRAFTSMAN CUNNING AT HIS TRADE



Table of Contents

Significant Research Conclusions 5

Copper Supply and China as the Marginal Producer 15

Copper Geology: Chile and the World's Deteriorating Deposits 29

Grade Is King 67

From Grade to Grunt and the Real Impact of Wage Inflation 101

Demand — Waiting for the Trend to Reassert Itself 121

Price Forecast 135

Appendix 145

Index of Exhibits 187



Exhibit 1 Financial Overview

Source: Bloomberg L.P., FactSet and Bernstein estimates and analysis.

9/20/2013 Anglo American BHP Billiton Glencore Xstrata Rio Tinto Vale MSCI Europe

AAL.LN BLT.LN GLEN.LN RIO.LN VALE3.BZ MSDLE 15

Rating O O O O O ‐

Local Currency Units £ £ £ £ BRL

Current Share Price (Local Currency) 13.95 18.45 2.70 29.54 31.31 387.48

52‐Week High (Local Currency) 20.72 22.36 3.98 37.57 44.10 390.06

Current Low (Local Currency) 12.07 16.67 2.57 25.82 28.39 317.35

YTD Performance (%) (17.9%) (4.2%) (6.0%) 2.0% (5.2%) 14.8%

YTD Relative Performance (%) (32.7%) (18.9%) (20.7%) (12.7%) (20.0%) ‐

12‐Month Price Target (Local Currency) 20.25 22.50 5.25 41.25 46.50 ‐

Potential Upside/(Downside) to TP 45% 22% 94% 40% 49% ‐

MCAP (US$m) 27,282 149,601 28,538 83,128 79,549 ‐

Net Debt/(Cash) (US$m) 9,719 32,169 22,551 21,769 26,088 ‐

Minorities 17,642 1,349 5,032 490 (4,702) ‐

EV (US$m) 53,688 174,499 95,458 103,030 100,935 ‐

EBITDA ‐ US$m Anglo American BHP Billiton Glencore Xstrata Rio Tinto Vale

2013E 10,002 25,614 15,902 24,183 22,523

2014E 11,494 38,119 20,203 31,348 26,230

2015E 14,005 45,903 26,252 38,698 32,415

EV/EBITDA ‐ US$m Anglo American BHP Billiton Glencore Xstrata Rio Tinto Vale

2013E 5.4 6.8 6.0 4.3 4.5

2014E 4.7 4.6 4.3 3.3 3.8

2015E 3.8 3.8 3.3 2.7 3.1

EPS ‐ US$/share Anglo American BHP Billiton Glencore Xstrata Rio Tinto Vale

2013E 2.10 2.70 0.43 5.27 2.28

2014E 2.64 3.73 0.63 7.19 2.55

2015E 3.10 4.73 0.97 9.33 3.58

Dividend Yield ‐ % Anglo American BHP Billiton Glencore Xstrata Rio Tinto Vale

2013E 4.0% 4.1% 3.4% 3.7% 2.2%

2014E 3.1% 3.4% 4.7% 3.8% 3.3%

2015E 3.4% 3.5% 5.6% 4.8% 4.6%



Significant Research Conclusions

Our analysis is complex, but the thesis is simple: We do not believe that the copper price has become detached from the fundamentals of supply and demand. We believe that the rising copper price over the last decade has triggered five effects critical for maintaining the supply/demand balance. First, it allowed for the entry of 1Mtpa of high-cost sub-scale material from China. Second, it enabled the exploitation of lower-grade sections in existing mines, thus offsetting the effects of the deposits' geological degradation. Third, it incentivized investment in new projects, despite the declining exploration success and increased geopolitical risk. Fourth, it accelerated the use of scrap copper as a substitute for mined material. Finally, it has witnessed the replacement of copper in nonessential applications with aluminum and steel. This Blackbook is primarily supply side focused and examines the first three of these effects. We see no reason to believe that the price agnosticism of consensus will hold true, and believe that copper will test the US$10,000/t mark by 2018. All the major miners have copper exposure but none more so than Glencore Xstrata, which stands to benefit the most, should our forecast hold true.

The consumption of commodities varies with their geological abundance. The world's economic system has adapted to use those materials that are most readily accessible. More interesting, though, is not the relationship between commodity abundance and its use itself, but rather the departures from this relationship. A low geological abundance and high use tells us that the metal in question is relatively more important for the world economy than a metal with a high abundance and low use. All other things being equal, a higher and stronger price should result for an overexploited metal and vice versa. In other words, incentivizing an increase in the supply of a metal with little availability and an already high rate of exploitation (absent improvements in the productivity of mining) requires prices to move upwards.

Exhibit 2 shows the relative overutilization and underutilization of the main industrial metals: The higher the figure, the more overexploited a commodity is. In this context, copper stands as the most overexploited metal, with a production 16.4 times greater than the underlying relationship between abundance and use would suggest. We already use copper at an incredible rate. Consequently, growing production necessarily involves accessing more challenging geology.

We Do Not Believe That the Copper Price Has Become Detached from the Fundamentals of Supply and Demand

Copper Is the Most Overutilized Major Commodity



Exhibit 2 Copper Stands Out as Being the Most Overutilized Commodity Relative to Underlying Geological Endowment — Testimony to the Industrial Importance of This Metal and the Difficulty in Growing Supply in Anything Other Than a Supportive Price Environment

Source: USGS and Bernstein analysis and estimates.

Copper porphyry deposits are critical to support global copper demand. Exhibit 3 looks at the history of copper porphyry discovery and the real copper price. Historically, copper mining was far more frequently an underground activity with human labor chasing rich seams. It was inherently much less productive and economic only insofar as labor was abundant and cheap.

The real price of copper halved in the 1920s, on the back of an explosion in copper porphyry exploitation. The demonstration of the economic viability of low-grade copper mining encouraged the delineation and discovery of significant number of very large deposits that had hitherto been thought un-mineable using human muscle power alone. In the post-World War II period, as the scale of new copper porphyries began to decline, the real price of copper began to rise. Against a rising demand environment (supported by electrification programs in the West), the technological and geological step-change in new porphyry discovery began to run its course.

An observer in the 1970s would have been forgiven for thinking that real price increases for copper were likely to continue indefinitely. However, two things intervened to change this — the step-change in Western copper demand growth after the oil shocks and the economic reforms in Chile. The Chilean economic reforms of the late 1970s and 1980s and the return of foreign investment saw a new wave of copper deposit discoveries. The supply of new material from Chile was sufficient to result in two decades of negative price performance for copper.

However, since then the rate of new large copper porphyry discoveries has ground to a halt. Moreover, demand has accelerated on the back of Chinese industrialization, and we are now back to the territory of real price increases. That is, we are back to where the copper price trend extrapolated from the end of the 1980s would have taken us, if the impact of Chile was taken out. It is worth noting that there is no new Chile on the horizon, and that Chile occupies a unique geological position as far as copper is concerned.

0

2

4

6

8

10

12

14

16

18

Cu Cr Pb Zn Fe Mo Ag Au P Mn Ni U Na K Al Co Pt Mg Pd Li V

Act

ua

l Use

Re

lati

ve to

Th

at I

mp

lie

d b

y G

eo

log

ical

A

bu

nd

an

ce

Over- and Underutilization of Geological Endowment by Commodity

Overutilized

Underutilized

Major New Discoveries Are Declining and Prices Are Rising



Exhibit 3 The Long-Term History of Copper Porphyry Discovery Had a Marked Impact on the Copper Price: Falling Prices Have Been Associated With Increased Finds of Relatively Few Massive Ore Bodies — We See No New Chile on the Horizon

Note: Bars are the average size of new copper porphyry discoveries and the line is the copper price.

Source: USGS and Bernstein analysis and estimates.

Chile's contribution to the supply of copper metal is hard to exaggerate (see Exhibit 4). It is 260% larger than its nearest rival, China. However, in terms of growth, the last decade has seen Chile stall and China emerge as the fastest-growing source of supply. The countries where growth has stalled (Chile, the U.S., Australia, Canada and CIS) contain all of the "easy" political locations for mining investment.

This is highly suggestive of the fact that the mining industry first invested in and developed the easiest projects from a risk-return perspective geology, which makes perfect sense. But it also serves as yet another reminder of the difficulty that future mined growth will encounter, as the copper industry is forced to take on ever higher risk to secure new sources of supply.

Exhibit 4 Chile Stands Out as the Most Important Source of Supply, But This Was Not Always the Case — While Chile's Growth Has Slowed, China's Has Accelerated

Source: Wood Mackenzie and Bernstein analysis and estimates.

3,000

4,000

5,000

6,000

7,000

8,000

9,000

10,000

0

2

4

6

8

10

12

14

16

18

Co

pp

er P

rice

-U

S$/

t -R

eal 2

011U

S$

Ave

rag

e S

ize

of N

ew C

u P

orp

hyr

y D

isco

very

-M

t Cu

0

1000

2000

3000

4000

5000

6000

Ch

ile

Ch

ina

Per

u

US

A

Aus

tral

ia

Zam

bia

Rus

sia

DR

C

Can

ada

Mex

ico

Min

ed C

op

per

Pro

du

ctio

n -k

t

2012 Mine Supply of Copper

-2%

0%

2%

4%

6%

8%

10%

Ch

ina

Per

u

Ch

ile

Aus

tral

ia

Mex

ico

DR

C

Zam

bia

Rus

sia

Can

ada

US

A

CA

GR

199

0-20

03

Mined Output Growth of Largest Copper Producers

China Is the Fastest-Growing Source and No. 2 Global Supplier of Copper



Copper price forecasting requires one to understand the industry's cost structure. We believe that today's copper supply structure and price setting are radically misunderstood by the broad market. Standard cost curves of Chinese copper supply lack approximately 1 million tons of new mined Chinese copper that have been added since 2003. We estimate the cost of these "missing mines" using data from the Chinese National Bureau of Statistics. Knowledge of the industry's margins in aggregate, copper price and total Chinese production enables us to estimate the structure of the residual part of the cost curve (see Exhibit 5).

China has been the fastest-growing source of new mined copper over the last decade, despite not having anywhere near the endowment of metal that would ordinarily lead one to expect such a feat was possible. No one really understands where this additional metal has come from and the nature of the mines that have been required to come on line to satisfy Chinese copper demand. There has been a huge growth in unknown and consequently "un-costed" metal over the last decade.

What we do know is that this growth in production has tracked the rise in copper price (see Exhibit 5). What this relationship indicates: It is only through the expedient of rising prices that supply has been able to match demand. Thus, we assert that it is a large number of small, low-grade and high-cost Chinese mines that stand at the right-hand side of the global cost curve and connect the cost curve to the price of copper.

Exhibit 5 A Corrected China Cost Curve Would Display Much More High-Cost Production: The

Relationship Between Price and Chinese Copper Output Tells Us That They Are Not Unrelated Phenomena

Source: Wood Mackenzie, NBS and Bernstein analysis and estimates.

A steep cost curve implies a price that is high relative to average industry costs, and thus a high margin for the low-cost producers. A flat cost curve implies a price that is low relative to average costs and, consequently, a low margin for the industry.

Of the three main cost drivers in mining, two — diesel and power — show very little geographical variation. This is as expected: oil, coal and gas markets trade, in large measure, on global markets with internationally determined prices. In contrast, there is very significant international variation in the third cost driver — the price of labor. Consequently, labor cost evolution has the greatest ability to drive differential mining cost escalation and so a margin-generating rotation in the cost curve.

12,000

11,000

10,000

9,000

8,000

7,000

6,000

5,000

4,000

3,000

2,000

1,000

0

2,0001,5001,0005000

Cumulative Production – kt Cu from Known and “unknown” Chinese Mines

C1 costs + SIB Capex –US$/t

R² = 0.8588

0

50

100

150

200

250

300

350

400

450

500

250 450 650 850 1050 1250

Co

pp

er P

rice

-U

Sc/

lb

Volume of "uncosted"Chinese copper - kt

The Missing 1Mt of Chinese Production Is High Cost

Chinese Supply Is Relatively Much More Costly — This Steepens the Cost Curve



Given that a multitude of low-grade and labor-intensive Chinese mines sit at the right-hand side of the global copper cost curve, the key determinant of future marginal cost — and hence price — will be the future evolution of these mines. Given the global nature of oil and energy price, a view on this evolution boils down to a view on the future development of the Chinese copper's labor costs.

Exhibit 6 shows copper mining costs in the U.S. compared to those in China. In real terms, we expect the U.S. mining costs to track roughly flat and China's to rise dramatically. Given that Chinese production already sits at the right-hand side of the global cost curve, this would lead to a steeper rather than flatter copper cost curve. Consequently, not only will the productivity of the global copper industry decline as the impacts of deeper mines and lower grades make themselves felt, but the costs associated with the lower productivity will also rise. As a result, in our view, the copper price must rise in real terms, thus generating further margin for the mining industry.

Exhibit 6 It Is the Differential Cost Escalation That Drives Real Commodity Price Increases in

USD Terms

Source: Wood Mackenzie, IHS and Bernstein analysis and estimates.

Total costs of a mine are determined by the amount of ore and waste moved and the volume of ore milled. The ore tonnage scales to the total cost base.

However, the grade determines the unit costs, as it determines the amount of metal available to bear the total costs of the operation. Consequently, the unit costs of a mining operation are inversely proportional to the grade of the material exploited. It is possible to show the sensitivity of operating costs to changes in mining parameters and efficiencies. To pick just one example, the hardness of a rock will determine the power that is required to grind that rock to a suitable size fraction. Exhibit 7 shows the huge differences in mining costs arise from differences in head grade. Copper reserves have only grown in the face of increasing consumption because the global reserve grade has collapsed. Reserves have held up only because resources have been converted into reserves by dramatically lowering the cut-off grade.

The inadequacy of reserves under current economic conditions necessitates changing the conditions that are deemed necessary to decide what counts as reserves. The geology is invariant, the economics must change. Even though the declines in global head grade (that is, the material that stands behind the cost

2.89%

6.95%

+140%

ChinaUSA

Differential Cost Escalation USA to China

Grade Declines Increase Costs Enormously — Grade Is the Most Important Determinant of Mining Costs



structure and price of copper today) may be well known, the implications of this for future copper prices are not.

What matters for the price are the relative declines of the prevailing head grade, the reserve grade and the cut-off grade. Critically, we see that 10 or so years ago, global head grades moved from being below global reserve grades to being significantly above them. The relationship between head grade and cut-off grade enables one to calculate the trajectory of future head grades. Because head grades are too high today, they must fall tomorrow. Our analysis would suggest that the risk of grades falling faster than many anticipate is high.

Exhibit 7 "Grade Is King" and Global Head Grades Are Declining

Source: Wood Mackenzie and Bernstein analysis and estimates.

Our analysis suggests the structural form for a multivariate regression that explains (without the introduction of specious exogenous regime shifts and Heaviside functions) the copper price evolution over the last 35 years.

The following regression explains nearly 90% of the variation in copper price over the last 35 years

Price = + 1 x Grade +2 x GDP +3 x Inventory These elements serve as a proxy for an economic determinant of price; in

particular, the cost structure of the industry (grade), global demand (GDP) and global supply (inventory). By far the most important driver of the three is grade. This relationship gives the effective real-world cost increase implied in grade decline (see Exhibit 8).

We can then use this relationship as the basis of a price forecast. We know that global grades are going to continue declining as currently sub-economic resources rather than the accelerated exploitation of existing high-grade reserves are called to satisfy demand growth. The IMF provides a useful global GDP forecast. Finally, we can assume relative stability in warehouse inventories. Even the most conservative grade profile gives rise to prices well in excess of US$10,000/t. It is only a question of when, not if, this occurs.

1,000

3,000

5,000

7,000

9,000

11,000

13,000

15,000

17,000

0.1

5%

0.2

5%

0.3

5%

0.4

5%

0.5

5%

0.6

5%

0.7

5%

0.8

5%

0.9

5%

1.0

5%

1.1

5%

US

$/t

Head Grade - %

Cost Sensitivity to Head Grade

0.60%

0.65%

0.70%

0.75%

0.80%

0.85%

0.90%

0.95%

1.00%

1980 1985 1990 1995 2000 2005 2010

Hea

d G

rad

e -

%

Global Copper Head Grade

Of the Movement in Copper, 90% Is Explained by Just Three Factors: Grade, GDP and Inventories



Exhibit 8 Nearly 90% of the Movement in Copper Is Explained by Just Three Factors: Grade, Global GDP Growth and LME Inventory

Source: Bernstein analysis and estimates.

In addition to focusing on the supply side, we must, of course, look at demand. The key point is that, recent weakness aside, we continue to believe that growth is nowhere near over. We define "copper intensity" as consumption per unit of output (i.e., copper per US$1,000 GDP), against the overall level of output (i.e., GDP per capita). We think of it as a "rate," showing how rapidly the country is embedding copper into the economy in a given period.

We define "cumulative copper intensity" or "copper capital stock" as the total historical copper installed per capita against the level of output. We look at this as a measure of the "level" of development, looking at how successful historic capital investment decisions have been.

We believe that both of these measures are necessary to gauge the current position of copper (and other commodities') consumption in an economy, as the term "rate" by itself only tells how fast an economy is developing. However, only by looking at the term "level" can we answer the question about the success and longevity of the industrialization. We believe that it is impossible to use the history of copper intensity alone to derive a forecast for future copper intensity.

Looking across the 35 countries (and country groupings) that actually consume the world's refined copper, we can chart how far down the path of industrialization each country is and the relationship of this to copper capital stock. It is only relatively late in a country's economic development that the bifurcation between a service- and a manufacturing-oriented economy actually takes place. There is a certain base load of metal that must be installed before this economic "choice" is made. In this context, it is important to note that China has a capital stock of copper equivalent to 29kg/capita — just 30% of Japanese levels and 22% of South Korean levels.

Knowing the end role of copper consumption provides the "missing" data point that is not supplied by an analysis of the rate of copper consumption alone. Consequently, the trajectory from the present into the future, in terms of copper intensity, can be constructed for any country.

The most important country for which we attempt this is China. We use the relationship between copper stock and output in Japan to model a demand-side trajectory for China. If we assume that copper productivity in China mirrors that

R² = 0.8864

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

10,000

0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 10,000

Fit

ted

Co

pp

er P

rice

-U

S$/

t

Actual Copper Price - US$/t

Copper Price Regression

We Use Copper Consumption and Capital Stock to Derive the Copper Intensity Curve for China



seen in Japan, the development of China's copper stock and its economy ought to follow the same pattern. This translates into the pattern for copper intensity development seen in Exhibit 9.

Exhibit 9 Our Analysis Gives Rise to the Following Trend Line for China's Copper Intensity

Source: Wood Mackenzie, USGS, Mitchell and Bernstein analysis and estimates.

As mining companies represent operationally and financially geared exposure to underlying commodity baskets (with ~80% of weekly equity price moves explained by moves in underlying commodity prices), we use a regression-based trading model and our forward commodity price forecasts to determine our 12-month price targets for our European metals and mining coverage.

To the extent that the regression holds, and the parameters of the regression have not significantly shifted, we take the target price from the trading model. To the extent that the regression is shifting or the equity is deviating, we look for evidence of whether this shift or deviation is temporary (and hence may be expected to close) or whether it signals a more fundamental re- or de-rating of the equity. In the event that there is no significant deviation or if we believe a deviation is temporary, the target price is set by the trading model. In the event that we believe a deviation is signaling a fundamental change, we will adjust our target price for this fundamental shift and disclose the manner and magnitude of the adjustment made. At present, no adjustments have been made to the target prices generated by our trading model. Note that we round final target prices in 25p/cent increments.

In addition to the target price (and short-term price forecasts generated by our trading model), we calculate a supplementary valuation that is DCF based. Given the long-lived nature of mining assets, we believe a DCF is critical to understanding the intrinsic value of a share (what the share price, in our view, "ought" to be today). Our DCF model is constructed in nominal local currency terms out to 2030, over which explicit commodity price and exchange rate forecasts apply. The nominal local currency cash flows are de-escalated into real USD cash flows and discounted at the company-specific WACC. A country risk premium reflecting the geographic origin of the cash flows is added to the underlying WACC to reflect cash flow items (i.e., expropriation) that cannot be explicitly modeled in the cash flow. All reserves are considered exploited by the model. In addition 50% of the incremental resources (i.e., 50% of the residual resources, excluding those

-

.10

.20

.30

.40

.50

.60

.70

.80

- 5.0 10.0 15.0 20.0 25.0 30.0 35.0

kg

/'000

$ G

DP

GDP/Capita (Real 2005$ PPP)

China Copper Intensity Forecast

China Actual China Trend China Forecast

Valuation Methodology



that have already been converted to reserves) of the company are modeled. Where residual life of the mine (LOM) may be inferred for operations beyond the 2030 time horizon, a terminal value is calculated for the remaining years of potentially exploitable material. We use this methodology to derive all forward-looking multiples and other valuation metrics. Note that we forecast our models in reporting currency (USD), convert to listing currency (British pound sterling or Brazilian real) at an average exchange rate, and round final DCF values in 25p/cent increments.

The four most significant risks facing the major mining houses are: 1) lack of capital discipline (specifically displacement of high-cost Chinese marginal producers by low-cost Western production), 2) operating cost inflation (USD-denominated unit costs in all the major mining houses have seen double-digit growth rates over the last 10 years, roughly half of which are macro related and the other half are real local currency), 3) a sustained downturn in the Chinese economy (the largest consumer of global resources) and 4) resource nationalism (ranging from increased share of rent extraction to outright asset confiscation).

Copper is the second-most important mined commodity after iron ore for the large miners, contributing 16% of EBITDA in 2012 (see Exhibit 10). Unlike iron ore, copper exposure is ubiquitous in our coverage group, with none of the world's largest mining companies lacking deposits of this commodity (see Exhibit 11). Glencore Xstrata is the most exposed, while Vale is the least. In particular, we like the decisiveness of Glencore Xstrata in its commitment to the development of the "new world" copper assets in Peru and Central Africa. For us, the most significant new copper development belongs to Rio Tinto, with Oyu Tolgoi in Mongolia. We also note that Rio Tinto has a number of longer-dated growth options in La Granja and, especially, Resolution. In addition, we remain "believers" in the copper price story. In our view, the fundamentals for copper are conducive to the price staying genuinely "stronger for longer" relative to, say, iron ore.

Exhibit 10 Copper Makes the Second-Most Important Contribution to the Cash Generation of the Miners

Exhibit 11 Glencore Xstrata Is the Most Highly Exposed in Our Coverage to the Copper Price and Vale Is the Least

Source: Corporate reports and Bernstein analysis. Source: Corporate reports and Bernstein analysis.

Iron Ore58%

Copper16%

Nickel1%

Zinc/Lead3%

Coal9%

Aluminum1%

Other12%

2012 Mining EBITDA (ex. Glencore trading) (100% = US$83bn)

38%

25%17%

9%3%

62%

75%83%

91%97%

Glencore Xstrata (100% = US$12bn)

Anglo (100% = US$9bn)

BHPB (100% = US$25bn)

Rio (100% = US$20bn)

Vale (100% = US$18bn)

Contribution of Copper to Mining EBITDA (2012)

Copper Other

Risks

Investment Conclusion



Rio Tinto (TP £41.25) — 91% of 2012 EBITDA from copper: We consider Rio Tinto the most attractive stock in our coverage. Rio also has exposure to some of the world's best operational copper assets — not to mention two of the world's best undeveloped copper deposits (Resolution and La Granja). Moreover, Rio owns some of the highest-quality iron ore assets and (vitally) infrastructure globally. Tier 1 assets in copper include Escondida, Grasberg, Bingham Canyon and Oyu Tolgoi, while iron ore includes Dampier and Cape Lambert (Australian Pilbara). The prospects of genuine capital discipline and cost cutting under new CEO Sam Walsh (who made his bones in low-cost brownfield Pilbara production), coupled with a more reasoned approach to volume growth, mean that we continue to see Rio Tinto as our top pick.

Vale (TP BRL 46.50) — 3% of 2012 EBITDA from copper: Vale is the world's largest iron ore producer and, in Carajas, has one of the most globally attractive iron ore assets. It is the most operationally geared miner to the iron ore price and we see an asymmetric risk to the upside in iron ore prices in the near term, hence from a pure value consideration, we are more favorably inclined to Vale now than previously. We do note the significant influence of the Brazilian government (5.5% directly through Golden Shares and ~34% indirectly through the strategic consortium of Valepar). The company is, in our view, most at risk should a reduced iron ore price eventuate, given its geographic distance from the world’s largest consumer of iron ore, China.

BHP (TP £22.50) — 17% of 2012 EBITDA from copper: BHP Billiton is our highest-quality stock. In 2012, the company generated the highest revenue of the "Big Three" miners in our coverage (US$72 billion versus US$56 billion for Rio and US$49 billion for Vale) and had the highest EBITDA margin (42% versus 39% for Vale and 36% for Rio). It is the second-most diversified (from a commodity exposure perspective) company in our coverage, after Anglo American. However, where platinum has been a drag on Anglo's portfolio, BHP's has received a boost from its petroleum division (69% EBITDA margins in calendar 2012 versus 11% for Anglo's platinum division). The strong commodity risk diversification is complemented with geographic diversification that, while skewed to Australia, is nonetheless low risk.

Anglo American (TP £20.25) — 25% of 2012 EBITDA from copper: Anglo American has the most diversified portfolio of our coverage group from a commodity perspective and is significantly smaller than the "Big Three" of BHP, Rio and Vale due to its smaller iron ore exposure. Kumba, however, like AmPlats and thermal coal, increases the company's country risk exposure due to its South African location and earns the company the highest country risk premium in our coverage — 2.6% versus an average of 1.2% for the "Big Three." Given recent operational difficulties (including failure to deliver on Minas Rio and Los Bronces, not to mention the issues plaguing platinum), Anglo remains in our minds a turnaround story — one that will challenge new CEO Mark Cutifani, but one we believe is solvable for a leader with his 36 years of operational expertise.

Glencore Xstrata (TP £5.25) — 38% of 2012 EBITDA from copper: Glencore Xstrata offers exposure for copper and coal bulls without a taste for iron. It also offers exposure to the sales and trading house that Mr. Ivan Glasenberg built. As such, it has a profile that is distinct relative to the pure miners in our coverage. The impact of the high-turnover, low-margin trading business (2% EBITDA margin in 2012 versus 25% for Glencore Xstrata combined) is clearer even when contextualized against the average 33% for our coverage group. As an owner operator (Mr. Glasenberg holds 8% of the combined entity's paper), the CEO's incentives are squarely aligned with investors. We do consider that Glencore has some ways to go yet before it will be "institutional quality" on the metrics of reporting and governance. Furthermore, the company's operations in frontier jurisdictions like the DRC not only result in the second-highest country risk premium in our coverage but also carry headline risk — as does Glencore's trading division.



Copper Supply and China as the Marginal Producer

In 1814, Thomas Cooper, in the Emporium of Arts and Sciences, stated "Copper is an article so necessary to us that I hardly know of any manufacture of such importance after iron." Wise words indeed, which, after a gap of 200 years, forecast exactly the relative importance of this metal to the earnings power of the miners — second only to iron ore. In the subsequent analysis, we lay out why we do not believe that the copper price has become detached from the mine supply cost curve over the last decade. As with iron ore, we trace the origin of structurally higher copper prices to the low levels of productivity in sub-scale domestic Chinese mined production. Moreover, we remain convinced that significantly higher copper prices will be required to continue to incentivize the new mine capacity required to clear demand in the face of increased geopolitical risk, falling grades and the declining success of exploration.

We should begin by pointing out that copper is the second-most important mined commodity after iron ore for large miners, contributing 16% of EBITDA in 2012 (see Exhibit 12). Unlike iron ore, copper exposure is ubiquitous in our coverage group with none of the world's largest mining companies lacking deposits of it (see Exhibit 14). Glencore Xstrata is the most exposed while Vale is the least. As we show in Exhibit 13, along with iron ore, copper has been the standout performer among the main industrial metals over the last decade. Despite these similarities, copper and iron ore look very different from the perspective of fundamental commodity economics.

As we will discuss in this Blackbook, the world's economic system has adapted to use those materials that are most readily available (see Exhibit 15). It is also worth noting that the economic role played by any metal depends upon its physical and chemical characteristics (e.g., mass to strength ratio), which make it hard to find substitutes for certain metals. Currently, copper stands out as one of the most overexploited metals with a production 16.4 times as great as the underlying relationship between abundance and use would suggest (iron ore at 7.4 times) — see Exhibit 16.

Cooper is not only more geologically challenging to mine in comparison to iron ore or aluminum, but its supply is heavily concentrated in a specific region. Chile is by far the most important producer of copper globally, and its mineral wealth is behind copper price being as low as it has been for the last two or three decades (see Exhibit 17). Chile's significance can be contextualized by simply noting that the country is much more important to global copper supply than Australia is to global iron ore supply. The scarcity of copper is evident in the fact that Australia has been able to deliver steady year-over-year growth in iron ore, while Chile's copper output has been stagnant.

We believe that Chile will have to work very hard just to maintain its copper output constant against a backdrop of falling grades and deteriorating geological quality.

As With Iron Ore, the Emergence of the Chinese Mines at the Margin of the Global Copper Cost Curve Drove the Fly-Up in Its Price Over the Last Decade

Each Commodity Respects Its Own Supply and Demand Fundamentals; a Key Difference Between Iron Ore and Copper Relates to the Difference in Geology Between Australia and Chile



Exhibit 12 Copper Is the Second-Most Important Contributor to the Cash Generation of the Miners…

Exhibit 13 …And Has Been Consistently Strong Over the "Super-Cycle"

Source: Corporate reports and Bernstein analysis. Source: Bloomberg L.P. and Bernstein analysis.

Exhibit 14 Glencore Xstrata Is the Most Highly Exposed in Our Coverage to the Copper Price While Vale Is the Least

Source: Corporate reports and Bernstein estimates and analysis.

Iron Ore58%

Copper16%

Nickel1%

Zinc/Lead3%

Coal9%

Aluminum1%

Other12%

2012 Mining EBITDA (ex. Glencore trading) (100% = US$83bn)

0

100

200

300

400

500

600

700

800

900

1000

Jun

e 20

00 =

100

Indexed Commodity Price Performance

Iron Ore Copper HCC

38%

25%

17%

9%3%

62%

75%

83%

91%97%

Glencore Xstrata (100% = US$12bn)

Anglo (100% = US$9bn) BHPB (100% = US$25bn) Rio (100% = US$20bn) Vale (100% = US$18bn)

Contribution of Copper to Mining EBITDA (2012)

Copper Other



Exhibit 15 There Is a Very Strong Relationship Between Geological Abundance and Industrial Use; the Most Useful Commodities, in Terms of Economic Application, Also Happen to Be the Most Geologically Abundant

Source: USGS and Bernstein estimates and analysis.

Exhibit 16 Copper Stands Out as the Most Overutilized Commodity Relative to Its Underlying Geological Endowment — a Testimony to the Industrial Importance of This Metal and the Difficulty in Growing Supply in Anything Other Than a Supportive Price Environment


Al

Fe

Mg

K

Na

PMn

V

Cr

Ni

Cu Zn

Co

Li

Pb

U

Mo

Ag

PdPt

Au

R² = 0.7852

0

1

10

100

1,000

10,000

100,000

1,000,000

10,000,000

1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05

An

nu

al P

rod

uc

tio

n (k

t)

Geological Abundance (ppm)

Use and Availability of Major Industrial Commodities

0

2

4

6

8

10

12

14

16

18


Act

ual

Use

Rel

ativ

e to

Th

at Im

pli

ed b

y G

eolo

gic

al

Ab

un

dan

ce


Overutilized

Underutilized



The Commodity ''Super-Cycle'' — The World Still Looks Very Different There has been a race to call the end of the nebulous phenomenon known as the "super-cycle," with every week a new straw man erected and demolished to prove the point. However, for us the world pre- and post-2003 continues to look radically different. We expect this view to persist until the 7 million workers employed in China's mines are replaced with capital.

As the readers of our work on iron ore1 will know, we believe that the regime change seen in commodity prices post-2003 has been driven by a fundamental change in the global cost curve structure for commodities (see Exhibit 18). More important than the extreme acceleration in commodity prices from 2003 to 2007 is the fact that the new price level has been sustained for the past five years. Whatever the "super-cycle" is, it has never been a belief that commodity prices would go on doubling indefinitely. However, many commentators still claim that commodity fundamentals do not drive commodity prices and that commodities have become "detached" from the normal dynamics of supply and demand. We disagree with this completely.

On the contrary, we believe that the microeconomics of the mining industry remain as they have always been. In this regard, commodity prices represent the interplay between capital and labor productivity (that is, capex and operating costs) geared through the underlying geological endowment. The evolution of this interplay results in the replacement of old and expensive mines with new and lower cost production sources as and when the price is deemed sufficiently high to justify new capital investment. To the extent that 2013 differs from 2003, it is the set of agents engaged in the commodity price formation rather than the nature of how

1 European Metals & Mining: Iron, Cold Iron, Is Master of Them All...or at Least 60% of EBITDA and European Metals & Mining: A Strange

Love — How I Learned to Stop Worrying and Love the Ore.

Exhibit 17 Chile Is Far More Important to Global Copper Supply Than Australia Is to Global Iron Ore; However, in Both Commodities, the Production from China (Despite the Poverty of Geological Endowment) Is Highly Significant

Source: Wood Mackenzie and Bernstein analysis.

The World Pre- and Post-2003 Continues to Look Radically Different, at Least Until the 7 Million Workers Employed in China's Mines Are Replaced With Capital

0

1000

2000

3000

4000

5000

6000

Chile China Peru USA Australia Zambia (& Northern

Rhodesia)

Russia Congo DR (Zaire & Bel

Congo)

Canada Mexico

Min

ed C

op

pe

r P

rod

uct

ion

-k

t




commodity prices are set that is different. Prior to 2004, the marginal unit of supply came from outside of China. Today it lies within it.

The Cause of Confusion Is Simply a Lack of Visibility Into China The root cause of the desire to believe that this time "something is different" in commodity pricing is the apparent breakdown of the simple pricing heuristic that prevailed for the last 30 years or so. Traditionally, the price of commodities was understood to be set by the cash costs of producing the marginal ton (typically taken as somewhere between the 80th and 95th percentile of the cash cost curve). However, from the middle of the last decade, this rule of thumb appeared to have stopped working (see Exhibit 19). Rather than despair about the validity of microeconomics, we note that for the heuristic to be of any use, two conditions need to be satisfied. The cost curve needs to be complete (i.e., cover all sources of supply). To the extent that the cost curve is not complete, it has to at least be

representative (i.e., those sources of supply that have been excluded should not be at or near the margin).

As shown in Exhibit 20, the first condition has not been met. Namely, for copper (just as for pretty much any commodity), visibility into large swathes of Chinese production is lacking, despite China being the second largest supplier of mined copper globally.

There has been huge growth in unknown and consequently "un-costed" metal supply over the last decade. What we do know is that this growth in production has simply tracked the rise in copper price (see Exhibit 21). This is more than just a

Exhibit 18 2004-06 Saw a "Fly-Up" in Commodity Prices, Which, When Compared to the Previous Cycles, Look Anomalous; We Strongly Believe That the Same Microeconomic Forces Are at Work Now as Previously and That Fundamentals, Rather Than "Funds," Offer the Best Explanation for the Movements in Commodity Prices

Source: IMF and Bernstein analysis.

China's Inability to Adapt Capital-Intensive Modes of Production Has Prompted the Change in the Position of the Chinese Mines on the Cost Curve; This Was Driven by the Increase in Chinese Wages in USD Terms, Making Chinese Mining Uncompetitive Globally

0

50

100

150

200

250

300

1980M01

1981M01

1982M01

1983M01

1984M01

1985M01

1986M01

1987M01

1988M01

1989M01

1990M01

1991M01

1992M01

1993M01

1994M01

1995M01

1996M01

1997M01

1998M01

1999M01

2000M01

2001M01

2002M01

2003M01

2004M01

2005M01

2006M01

2007M01

2008M01

2009M01

2010M01

2011M01

2012M01

2013M01

IMF Metals Price

Index ‐2005 = 100

Key features of the behaviour of commodity prices prior to 2005 The time from trough to peak of any cycle was typically 1 year

From peak to trough, prices declined along a smooth monotonic trajectory towards a fixed and "well known" long term price The fundamental question for miners was not why but how to execute the next mine, essentially a

question for engineering not economics as cost was more important than price. The key question now facing the miners is "how much Western capital ought to be expended to

underwrite the industrialisation of China?" this is not an engineering problem and requires a fresh approach to how the miners think about value.

The duration of any price spike was

limited, lasting no more than two years

Price paths post a price spike were smoothly

downward trending

Prior to 2005 long term price had a clear meaning

tied to the floor towards which prices reverted.

Post 2005 it was clear thatthe dynamic had changed,

the key question is why? WIthout this

understanding mining value is impossible.



coincidence. Once again, it points to the fact that it is not prices that have become detached from the cost curve, but that it is only through the expedient of rising prices that supply has been able to match demand.

In our view, low geological quality of China's endowment with certain commodities (copper, coal [in part] and iron ore) makes capitalization of the mining industry difficult. Consequently, it remains very labor intensive: the use of significant manpower is the norm in China, while in the West human labor has been largely displaced by machinery. As long as mining wage rates in China were low by international standards, this did not matter. However, once labor became expensive in USD terms, Chinese mining became uncompetitive globally. This led to a rotation of the cost curve in a manner not seen before due to the lack of data. In other words, we moved from a regime where the "known" cost curve was incomplete, yet representative of the total — i.e., where Chinese domestic mines were on the left-hand side of the global cost curve — to a regime where the "known" cost curve was incomplete and unrepresentative of the total — i.e., where the Chinese mines were now on the right-hand side of the cost curve. The rapidity of the shift of Chinese mines along the cost curve mirrored the pace of China's overall industrialization. Thus, the apparent change in the regime from 2004 was not due to the collapse of microeconomic principles, but simply due to our inadequate understanding of the marginal source of supply.

Exhibit 19 At First Glance, the Old Heuristic of Marginal Cash Costs Appeared to Have Broken Down…

Exhibit 20 …But a Better Answer Is Provided by the Negligible Visibility Into China's Mining Costs

Source: Wood Mackenzie, Bloomberg L.P. and Bernstein analysis. Source: Wood Mackenzie, Bloomberg L.P. and Bernstein analysis.

0

50

100

150

200

250

300

350

400

450

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

US

c/lb

—N

om

inal

Copper Costs and Price

C(90) Price

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Chile China

% o

f Min

e S

up

ply

fro

m W

ho

se C

ost

s A

re K

no

wn

% of "Costed" Copper Supply

Known Unknown



Exhibit 21 Virtually Nothing Is Known About the Cost Structure of Chinese Copper Mining (as Shown in the Previous Exhibit), Yet the Relationship Between Price and Chinese Copper Output Tells Us That They Are Not Unrelated Phenomena

Source: NBS and Bernstein estimates and analysis.

Reconstructing the Missing Chinese Mines We aim to complete the unknown part of the global cost curve by estimating the supply coming from the missing Chinese mines and plugging it back into the cost curve. Without an understanding of the industry cost structure, forecasting copper price is impossible.

As with iron ore, we believe that the present copper supply structure and price setting are radically misunderstood by the broad market. In this regard, Exhibit 22 shows what is known with reasonable certainty about the highest-quality portion of the Chinese copper industry (which is also not accidentally the most visible portion). However, this cost curve lacks approximately 1 million tons of new mined Chinese copper that has been added since 2003. We estimate the cost of these missing mines using data from the Chinese National Bureau of Statistics (NBS), which reports on the aggregate industry profitability in China. Knowledge of the industry's margins in aggregate, copper price and the total Chinese copper production enables us to estimate the structure of the residual part of the cost curve (see Exhibit 23). In essence, whatever margin the industry is currently generating, it is attributable to the low(ish) cost mines shown in Exhibit 22. Exhibit 24 gives the details of this calculation. The remaining production is unevenly distributed towards the right-hand side of the cost curve. Consequently, it acts to support copper price.

The completed Chinese cost curve can be combined with the reasonably well-known cost curve for the remainder of the global copper industry. Exhibit 25 and Exhibit 26 show the global cost curve, excluding and including this high-cost material. While the difference between the two curves appears marginal, the effect is an important one. By increasing the volume of material on the right-hand side of the cost curve, we are able to refute the thesis that prices have become detached from the cost curve. In addition, we see additional price support for copper in the medium to longer term. While we believe that the cash cost curve for any commodity gives the price of that commodity over any meaningful period (i.e., most commodities follow an economically efficacious price discovery mechanism), the level of the inflection of the cost curve provides the cash flow and the incentive for new production capacity. It is the interaction of the current cash cost curve and the current incentive price curve that generates the future cash cost curve and, with

R² = 0.8588

0

50

100

150

200

250

300

350

400

450

500

250 450 650 850 1050 1250

Co

pp

er

Pri

ce

—U

Sc/

lb

Volume of "Uncosted" Chinese Copper — kt

A Completed Chinese Cost Curve, Obtained by Estimating the Residual Part, Allows Us to Refute the Thesis That Prices Have Become Detached from the Cost Curve; This Also Indicates Price Support in the Medium to Longer Term



it, the future price for a commodity. The rapidity with which new assets are brought on line and the place that they occupy in the industry's cash structure determine the evolution of price over time.

Exhibit 22 The Apparent Cost Structure of the Copper Industry in China Is Based on a Highly Non-Representative Selection of Mines; Moreover, It Fails to Adequately Describe the Profitability of This Sector

Exhibit 23 A Corrected Chinese Cost Curve Would Display Much More High-Cost Production

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie, NBS and Bernstein estimates and analysis.

Exhibit 24 Calculating the Volume and Costs of the Missing Portion of the Chinese Data Is Possible Using Data from the NBS

Source: NBS, Bloomberg L.P. and Bernstein estimates and analysis.

11,000

4,000

3,000

2,000

1,000

0

450350 40050 100 150 200

12,000

0

10,000

9,000

8,000

7,000

250 300

6,000

5,000

Cumulative Production – kt Cu from Known Chinese Mines


12,000

11,000

10,000

9,000

8,000

7,000

6,000

5,000

4,000

3,000

2,000

1,000

0

2,0001,5001,0005000

Cumulative Production – kt Cu from Known and “unknown” Chinese Mines


Completing the Chinese Cost Curve

2012 Average Cu Price US$/t 7,949 Bloomberg

2011 Chinese Non‐Ferrous Profit Margin % 16.5% NBS

Average Cash Costs of Know Mines US$/t 3,898 Calculation

2012 Profit Margin of Known Mines % 25.5% Calculation

2012 Total Domestic Cu US$/t 1,542 Kt, WM

2012 Identified Cu kt 435 Calculation

2012 Residual Cu US$/t 1,107 Calculation

2012 Margin of Residual Mines % 13.0% Calculation

Average Cash Costs of Residual US$/t 5,879 Calculation

Minimum Cash Costs of Residual US$/t 4,183 Calculation

Maximum Cash Costs of Residual US$/t 7,925 Calculation



Exhibit 25 This Mined Tonnage Needs to Be Factored Into the Global Cost Curve...

Exhibit 26 ...Where It Is Concentrated Towards Right-Hand Side of the Curve, Filling Out the High-Cost Volume

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis.

Chinese Labor Productivity and Cost — The True Cause of the ''Super-Cycle'' Following the economic reforms of the 1980s and 1990s, China appeared to be embarking on rationalization and consolidation of its mining industry. However, this trend came to an abrupt end in 2003 (see Exhibit 27). Concurrently, mining wages started to follow the exponential growth rate seen in the economy as a whole (see Exhibit 28).

In addition, from 2003 onwards, the mining industry experienced an above-average wage growth, thus suggesting a requirement to pull labor into the sector away from other, perhaps more attractive, employment opportunities that industrializing economies can offer. Taken in RMB terms, this led to a 24% CAGR in the costs attributable to labor alone. Given a gradual appreciation in RMB, this led to a 28% CAGR in USD terms (see Exhibit 29).

From 2003 to present, the Chinese output of mined commodities has increased significantly, particularly in iron ore (see Exhibit 30). In aggregate, almost continual gains have been recorded in labor productivity on a copper equivalent basis since 1990 (see Exhibit 31). However, since 2003, China's mining labor productivity started to decelerate at the same time as mining wages started to accelerate (see Exhibit 32). While historically labor costs and productivity in Chinese mining tracked each other (as one would expect them to do), 2003 marked a breakdown of this relationship. Subsequently, wages in Chinese mining were driven not by output from mining, but rather by productivity in other sectors (e.g., manufacturing).

It is also worth stressing the sheer scale of the Chinese mining industry in terms of employment. In aggregate, nearly 7 million people are employed in mining in China, compared to 42,000 employed in the U.S. (0.9% of the Chinese labor force versus 0.03% of the U.S. labor force — a factor of 33 times greater). Clearly, such labor intensity is possible only when labor is both abundant and cheap. Given China's demographics and the stage of economic development, both of these conditions have started to come under pressure. We believe that it will only become more acute as China continues to urbanize.

We strongly believe that this dynamic is responsible for the rise of the commodity "super-cycle." As such, the fly-up in commodity prices is explained entirely by the traditional microeconomics of the mining industry. Exhibit 33

3,000

10,000

5,000

8,000

4,000

2,000

0

12,000

1,000

6,000

9,000

11,000

15,000 20,000

7,000

0 5,000 10,000


Global Cumulative Production – kt Cu Including Known Chinese Mines

10,000

7,000

3,000

4,000

10,000

5,000

1,000

6,000

12,000

8,000

0 5,000 20,000

11,000

9,000

2,000

15,000

0


Global Cumulative Production – kt Cu Including Known and “unknown” Chinese Mines

Nearly 7 Million People Are Employed in the Chinese Mining Sector; the Abundant and Cheap Labor That Has Allowed China to Maintain High Labor Intensity Has Come Under Pressure, and We Believe It Will Remain So as China Continues to Urbanize



shows the rise in Chinese mined labor costs against the rise in aggregate commodity prices. Exhibit 34 displays that the simple expedient of adding global growth rates to capture cyclicality explains almost the entirety of the observed data (R-squared = 96%).

In essence, China has moved from a low productivity, but very low-cost mining location (and thus in aggregate being on the left-hand side of the global mining cost curve) to a still low productivity, but now relatively high-cost location (hence, currently on the right-hand side of the global cost curve).

Exhibit 27 Chinese Mining Employment Started to Increase Slowly Post 2000...

Exhibit 28 ...With Mining Wages Tracking Exponential Growth Seen in the Economy as a Whole

Source: NBS and Bernstein analysis. Source: NBS and Bernstein analysis.

Exhibit 29 In USD Terms, Labor Costs Have Grown in Mining at a CAGR of 28% Since 2003…

Exhibit 30 …Which Is Significantly Above the Growth Rates in Domestic Mining Output


0

100

200

300

400

500

600

700

800

900

1000

Chinese Mining Employment

'0,000 People

0

10,000

20,000

30,000

40,000

50,000

60,000

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

2010

Lab

or W

ages

—Y

uan

per

an

nu

m

Chinese Wages

Mining Total

0

10,000

20,000

30,000

40,000

50,000

60,000

US

$ -M

illio

n

Total Chinese Mining Labor Cost

13%

6%6%

7%

8%

5%

Iron ore Coal Copper Zinc Nickel Bauxite

Growth of Mined Commodities in China

Output CAGR 1990-2011



Exhibit 31 Chinese Mining Productivity Started to Decelerate from 2003…

Exhibit 32 …And a Fundamental Disconnect Appeared Between Mining Wages and Output


Exhibit 33 The Exponential Rise in Chinese Mining Labor Costs Has Driven a Corresponding Rise in Commodity Prices

Exhibit 34 If the Cyclical Impact of Global GDP Is Added, an R-Squared of 96% Is Returned

Source: IMF and Bernstein analysis. Source: IMF and Bernstein analysis.

Clearly, the rise in Chinese labor costs would not have mattered to global commodity prices, if China's output of commodities had been able to track demand, and if China remained self-sufficient in its consumption of raw materials. However, the failure of productivity resulted in the need to import raw materials on a massive scale. As soon as hitherto disconnected global commodity markets started interacting with the domestic Chinese markets, the price of imported commodities rose until the Chinese consumers were ambivalent between sourcing the required material domestically or through international markets. The reason behind the

0

1

2

3

4

5

6

7

8

9

Cu

Eq

uiv

alen

t Ou

tput

per

Wo

rker

(t)

Chinese Mined Output per Worker

9%

4%

9%

22%

1990-2003 2003-2011

CA

GR

Chinese Mining Wages vs. Productivity Growth

Productivity Wages

0

50

100

150

200

250

300

350

400

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

Chinese Labor Costs vs. Commodity Prices

IMF Metals Index Chinese Mining Labor Costs

0

50

100

150

200

250

300

350

400

Chinese Costs and Global Growth vs. Price

IMF Metals Index Chinese Costs + GDP

The Reason Behind the Massive Increase in Raw Material Imports Into China Despite the Accelerating Cost of Those Materials Is Simply That It Was the Lowest-Cost Alternative



massive increase in raw material imports into China despite the accelerating costs of those raw materials is simply that it was still the lowest cost alternative.

The pertinent questions thus become: how long will this state of play persist? Why, if mining is a mature industry, cannot costs and labor be driven out of the Chinese industry? This is the age-old equation in mining between capital and labor productivity. The questions being asked come down to understanding why China has not transitioned to capital-intensive modes of metal production. We believe that there are three underlying reasons why this has not happened thus far.

Why Does China Not Transition to Capital-Intensive Production? There are three reasons why China has not transitioned to capital-intensive modes of production.

Geological endowment: Capitalization of mining assets is (in large measure) a move to ever more massive scale. However, this scale means that the life of mine (LOM) of assets will be proportionally reduced for a given geological endowment. Thus, capitalizing assets is economically feasible only where a corresponding scale exists in the underlying geology. We do not believe that such scalable assets exist in China in sufficient measure, particularly in the core commodities of iron ore and copper.

Efficiency of capital allocation processes: Even if the geological endowment were present in China, the process of moving away from labor requires large amounts of capital investment in relatively few assets and projects. This requires a depth of experience in capital allocation that the relatively immature capital markets and corporate governance structures in China do not provide. The mute witness to this is the proliferation of operations that have sprung up in the mining sector over the last 10 years.

Political will: Finally, even if the geological endowment were present and the capital allocation mechanisms existed to be able to identify the most economically productive path forward, the transition would require the displacement of over 6 million workers. While other forms of employment may be present in a rapidly growing economy, the fear is always that the transition will result in wholesale unemployment. This process could be incredibly politically disruptive (witness the miner strikes in the U.K. — a country with an established and generous welfare system).

We believe that all three of these requirements are either lacking or limited in China today. We believe that Chinese mining costs will track ever higher until such a point that the requirement for Chinese mining labor is obviated. We also believe that this will only happen when mining output from other regions with greater geological endowment and/or lower wages increases significantly (i.e., Africa, Latin America and Australia). Until that happens, there is asymmetric risk to the upside for commodity prices. Against a normalizing demand environment, any pull on the global commodity supply chain will see the copper price rally sharply. Only when we see the "rebalancing" of the Chinese mining industry, will there be the true end of the "super-cycle" with structurally lower commodity prices over a sustained period.

We believe that there is a paradox at the heart of the Chinese economy — preindustrial modes of natural resource production are attempting to coexist with industrial modes of production in other sectors. This has led to a decoupling between the costs and the productivity of Chinese mining, which we see are driving up both the imports of commodities into China and the costs of commodities globally. We further believe that such a situation will last until either China can access alternative non-Chinese mining assets (e.g., in Africa), which it can use to displace domestic mines, or Western mining companies do so on behalf of China. Thus, the duration of the "super-cycle" is largely in the hands of the current mining incumbents and comes down to the question of how much Western capital will be

China Has Not Transitioned to Capital-Intensive Modes of Production Due to a Lack of Geological Endowment, a Lack of Experience in Capital Allocation Required for the Initial Capital Investments, and a Lack of Political Will to Displace Several Million Workers from the Chinese Mining Sector

The Duration of the ''Super-Cycle'' Is in the Hands of the Current Mining Incumbents; It Boils Down to the Question of How Much Western Capital Will Be Put at Risk to Underwrite the Industrialization of China



put at risk to underwrite the continued industrialization of China (see Exhibit 35 and Exhibit 36).

Exhibit 35 Coal Mining in the U.K. in 1840… Exhibit 36 …And in China Today

Source: Royal Commission for the Employment of Children in Mines 1842.

Source: EPA.



Copper Geology: Chile and the World's Deteriorating Deposits

The two critical questions this chapter attempts to answer are "what are the consequences of Chile's unique geological endowment?" and "are current copper reserves sufficient to satisfy future copper demand?" Our analysis suggests that no other supply location has the potential to mirror the history of the Chilean copper growth. Hopes that either the Democratic Republic of Congo (DRC), Zambia or Peru will provide the world with another period of abundant copper supply at low prices are inconsistent with the underlying geology of those locations. We also explain how the financial implications of copper mine development link a country's underlying geological endowment to its eventual maximum mined production. Consequently, we are able to determine that existing reserves will be sufficient to meet only one-third of the world's incremental copper demand by 2030. This issue gets compounded by the 20+ year lead time between new copper discoveries and eventual exploitation, which renders yet undiscovered deposits insufficient to make up for the shortfall (even if there was evidence of ongoing copper exploration success, which is not the case). Rather, new copper supply will have to come from the already known sources that are uneconomic to develop at today's prices. The implication of this is clear — new copper supply is predicated upon higher prices than those that prevail today.

Overexploited Copper and the World's Deteriorating Deposits The geology and chemistry of metals within the Earth's crust form the basis of mining, and hence all the economic activity of the modern industrial society. While over 3,000 different minerals have been identified, only 30 of them form most rocks in the planet's crust. The list of chemicals on which these common rocks are based is even shorter and dominated by just nine elements: oxygen, silicon, aluminum, iron, calcium, magnesium, sodium, potassium and titanium. In fact, these nine elements account for 99% of the Earth's mass (see Exhibit 37).

Most commonly encountered minerals are silicates (of one form or another) interspaced with oxides, hydroxides and carbonates. These elements collectively constitute a category known as the "geologically abundant elements," which consists of aluminum, iron, magnesium, potassium, titanium and manganese. The economics of these metals are not primarily a function of their geological grade, and it is fair to say that we will never "run out" of them.2 However, this class of elements does not include copper — no rock-forming mineral contains copper as an essential chemical consistent.

2 Just to make this clear, the prices of titanium and aluminum relate primarily to the price of the power required to strip oxygen from the metal.

For iron ore, it relates to the capital tied up in mass logistics systems in general and deep-water port capacity in particular.

What Are the Consequences of Chile's Unique Geological Endowment? Are Current Copper Reserves Sufficient to Satisfy Future Copper Demand?

Once Exploited a Deposit Is Gone Forever, and Once Discovered a Deposit Cannot Be Rediscovered; the Earth's Endowment With High-Grade Copper Deposits Is Finite, and the Question Is When (Not If) That Limitation Will Lead to Higher Real Copper Prices



Exhibit 37 Nine Elements Account for 99% of the Earth's Mass; Iron and Aluminum Are Among Them, While Copper's Abundance Is Radically Different

Source: Pearson and Bernstein estimates and analysis.

There is a very strong correspondence between a metal's geological abundance and its use — this is unsurprising. What is surprising is the departures from this relationship, as they tell us something about the relative demand and relative importance of a commodity. In this respect, copper is unique.

Exhibit 38 shows how the consumption of major industrial commodities varies with their geological abundance (we show this on a log-log basis where several orders of magnitude of variation can be compressed). Clearly, the world's economic system has adapted to use those materials that are the most readily accessible. However, it is also worth noting that the economic role played by any metal depends upon its physical and chemical characteristics (for example, mass to strength ratio), which are not easily substitutable. The high geological abundance of iron in the Earth's crust stands as one of the key forces behind the industrial development of mankind. Even the use of aluminum — the closest metallic industrial substitute for steel and the closest element in terms of geological abundance — is predicated upon prior development of steel. Without the metallurgical properties of steel, the process of electrification and power generation necessary to develop aluminum as an economic metal would not have taken place.

Perhaps more interesting is not the relationship between commodity abundance and its use in itself, but rather the departures from this relationship. For example, low geological abundance yet high use would tell us that the metal in question is more important to the global economy than a metal with high abundance and low use.3 Consequently, all other things being equal, a higher and stronger price should result in an overexploited metal and vice versa. In other words, incentivizing an increase in the supply of a metal with little availability and an already high rate of exploitation (absent improvements in mining productivity) requires prices to move upwards. In contrast, for metals whose exploitation is low relative to abundance,

3 Of course, geological abundance is not the same as a metal's reserves or resources, which add an economic filter to the underlying geological

endowment. In order to be economically accessible, all elements require some further geological process of enrichment and concentration. So, while iron ore has an average abundance of 5.6%, it requires an approximately tenfold concentration of that abundance to generate the 60% Fe grade ores exploited in Australia and Brazil. Likewise, copper has a geological abundance of 0.006%, thus requiring an approximately hundredfold concentration of 0.6% to be economically viable.

Oxygen, 45%

Silicon, 27%

Aluminium, 8%

Iron, 6%

Calcium, 5%

Magnesium, 3%Sodium, 2%

Potassium, 2% Titanium, 1%

Others, 1%

The Elements of the Earth's Crust

Copper Occupies a Unique Role in the Industrial Society; It Is the Most Overexploited of the Main Metals, Testifying to Its Industrial Importance



any increase in price will simply trigger a wave of new supply, making sustained high prices harder to achieve.

In Exhibit 39, we show the relative overexploitation of the main industrial metals. The higher the figure, the more overexploited a commodity is. In this context, copper stands out as the most overexploited metal with a production of 16.4 times as great as the underlying relationship between abundance and use would suggest. Likewise, iron ore is 7.4 times overexploited and aluminum is significantly underexploited. This relative exploitation suggests that despite its high geological abundance, growing iron ore production is not as trivial a matter as many seem to think. This is even truer for copper. We already use these metals at incredible amounts. Consequently, growing their production necessarily involves accessing more challenging geology.

Exhibit 38 A Very Strong Relationship Exists Between a Commodity's Geological Abundance

and Its Industrial Use; the Most Useful Commodities in Terms of Economic Application Also Happen to Be the Most Geologically Abundant

Source: USGS and Bernstein estimates and analysis. Exhibit 39 Copper Stands Out as the Most Overutilized Commodity Relative to Its Underlying

Geological Endowment; This Testifies to the Industrial Importance of This Metal and the Difficulty in Growing Supply in Anything Other Than a Supportive Price Environment


Al

Fe

Mg

KNa

PMn

V

Cr

Ni

Cu Zn

CoLi

Pb

U

Mo

Ag

PdPt

Au

R² = 0.7852

0

1

10

100

1,000

10,000

100,000

1,000,000

10,000,000

1.E-03 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05

An

nu

al P

rod

uct

ion

(kt)

Geological Abundance (ppm)

Use and Availability of Major Industrial Commodities

0

2

4

6

8

10

12

14

16

18


Actu

al

Use

Rela

tiv

e t

o T

hat

Imp

lied

by

Geo

log

ica

l A

bu

nd

an

ce


Overutilized

Underutilized



Copper (and a number of other important metals) are classified as geologically scarce. It is within this metal class that geological grade plays the decisive role in mining economics. As we mentioned earlier, copper does not form part of the chemical composition of any common rock-forming mineral. Rather, copper exists as individual atoms substituting for other metals at an atomic level within a host rock. For example, copper will substitute out iron in the silicate mineral pyroxene (FeSiO3), without altering the fundamental nature of the mineral in bulk. However, there is a fundamental limit to the solubility of copper in any solid solution. The evidence suggests that this limit is much lower than the lowest grades of any copper ore that has ever been encountered. When we observe a rock with a grade of above ~0.1% copper (Cu), we actually see a copper-rich sulphide (or oxide) mineral with a grade of 30%+ disseminated through a copper barren host rock, thus giving rise to the average 0.1% grade (see Exhibit 43). However, no common rock has ever been encountered (and the sampling of common rocks is pretty exhaustive) with a grade of copper as high as 0.1%. Furthermore, the crustal abundance of copper is somewhere close to 0.007%. This implies that there is a sharp discontinuity between copper ores, wherein copper exists in a concentrated sulphide or oxide form and where copper exists dissolved within a silicate matrix. As copper concentrations approach their saturation levels in the silicate host, there is the generation of a new material form rather than any "super saturation" effect. Put another way, the highest copper grades observed in common rock do not overlap with the lowest grades observed in ores. Now, there is an order of magnitude difference in the energy required to liberate copper metal from a silicate rather than a sulphide. This then gives rise to the famous mineralogical barrier for copper (see Exhibit 40). We cannot go on indefinitely dropping the grade of mined copper — there is a hard stop at grades approaching 0.1%, at which there is a radical rather than continuous change in the cost structure of mined copper.

In addition, this suggests the equally famous bimodality of copper ore distributions versus the unimodality that is more typically associated with the geologically abundant metals (see Exhibit 41 and Exhibit 42).

Exhibit 40 Below Copper Grade of 0.1%, a Step-Change Emerges in the Cost Structure of

Copper Extraction; Consequently, We Face a Hard Stop in Our Ability to Exploit This Metal, Once High-Grade Deposits Are Exhausted

Source: Bernstein estimates and analysis.

1.00E+07

1.00E+08

1.00E+09

1.00E+10

1.00E+11

0.001% 0.010% 0.100% 1.000% 10.000%

En

erg

y R

equ

ired

To

Gen

erat

e C

op

per

-J/

kg

Copper Grade

The Mineralogical Barrier for Copper

The mineralogical barrier...below 0.1%

grade a step change in cost structure.

For Geologically Scarce Metals, Grade Plays the Decisive Role



Exhibit 41 Under a Unimodal Distribution, Grade and Tonnage Are Continuous

Exhibit 42 Under a Bimodal Distribution, Grade and Tonnage Are Discontinuous, With Clear Implications for Cost and Availability of New Material

Source: Bernstein estimates and analysis. Source: Bernstein estimates and analysis. Exhibit 43 Copper Ores Represent the Dissemination of High-Grade Copper-Bearing Minerals

Within a Barren Matrix; Copper Mining Is the Process That Separates These Valuable Minerals from the Worthless Gangue

Source: Wood Mackenzie, Gordon and Bernstein estimates and analysis.

Lo

g T

on

nag

e

Log Grade

Unimodal Grade Distribution

Mineralogical Barrier

Current exploitation

Lo

g T

on

nag

e

Log Grade

Bimodal Grade Distribution

Current exploitation

Mineralogical barrier

Mineral Chemical Composition % Cu by Mass Ore Type

Cuprite Cu2O 89 Oxide

Tenorite CuO 80 Oxide

Atacamite Cu2Cl(OH)3 60 Oxide

Malachite Cu2O(OH)2CO3 58 Oxide

Azurite Cu3(OH)2(CO3)2 55 Oxide

Chrysocolla Cu2H2OSiO3 36 Oxide

Chalcocite Cu2S 80 Secondary Sulphide

Covellite CuS 67 Secondary Sulphide

Bornite Cu5FeS4 63 Primary Sulphide

Digenite Cu9S5 78 Primary Sulphide

Enargite Cu3AsS4 48 Primary Sulphide

Tetrahedrite Cu12Sb4S13 58 Primary Sulphide

Chalcopyrite CuFeS2 35 Primary Sulphide



An Overview of the Copper Mining Process Value maximization in many mining processes involves scaling of capital equipment to the underlying geology. Consequently, efficient low-cost copper production is predicated upon the existence of massive high-quality ore bodies. Remove this feature and mining costs rise exponentially.

Mining is essentially a process of efficient material movement in pursuit of separating valuable from valueless minerals. There are two basic mechanisms at work by which this separation is achieved. The first one occurs at the mine site, where waste material is separated from the valuable ore. The second one occurs at the processing site, where the valuable metal-containing mineral in the ore is separated from the worthless host rock or gangue. Exhibit 44 presents a schematic illustration of this process.

Efficient Copper Mining Involves Bulk Material Movement, Which Requires a High Degree of Capitalization and a Corresponding Quality and Scale in the Underlying Geology

Exhibit 44 Copper Mining Involves the Identification, Liberation and Sale of Copper-Bearing Minerals; This Is Achieved Through Two Processes of Waste Removal; the First One Occurs at the Mine Site Where Ore Is Separated from Waste; the Second One Happens at the Milling/Flotation Site Where Concentrate Is Separated from Tailings




Within the mining step, the most important division is between underground and open-pit mining methods. The main advantage of underground mining is the selectivity of the mining method, which enables one to focus extraction on high-grade mineralized zones, while leaving waste and low-grade ore in situ. However, this selectivity comes at a cost. Underground mines must be ventilated and dewatered. Blasting is necessarily rather small scale, and the process of hauling material to the surface is very energy intensive. By contrast, open-pit mining does not have the expense of ventilation, while dewatering costs are often a fraction of those incurred in underground operations. Moreover, the process lends itself to economies of scale, with efficiencies gained through the utilization of ever larger haul trucks and mining shovels as well as through the use of scalable blasting programs.

The difference in the cost structures of these two methods is evident in the amount of energy required to move a ton of rock. While this varies considerably from operation to operation, the energy in an underground mine may be ~300MJ per ton versus less than 40MJ per ton in an open pit mine. Clearly, the open-pit mining methods lend themselves to the development of massive low-grade copper deposits such as are currently the mainstay of the world's copper production.

Within the mined supply of copper, the most important division occurs between the conventional milling route and what is termed SxEw (or solvent extraction and electrowinning). The milling route accounts for ~80% of the current mine supply. It seeks to exploit sulphide copper minerals. The basic steps of this process are laid out below and illustrated schematically in Exhibit 45. Mining: This involves separating ore containing the valuable metal-bearing

mineral from waste rock. The distinction between valuable and worthless rock is achieved via the cut-off grade, which delineates the minimum amounts of metal that a volume of material needs to contain to render its further treatment economical.

Comminution: This is the crushing and grinding of the ore in order to achieve physical liberation of the particles containing valuable mineral from the gangue matrix of worthless material in which those particles reside.

Beneficiation: This involves separation of the particles liberated by comminution in order to maximize the resulting concentration of the valuable mineral. For copper, this is achieved through the process of froth flotation. A solution consisting of the ground ore, water and a mix of various chemical reagents is created. These reagents bind preferentially to the surface of copper containing sulphides, so that when the solution is agitated, these particles float to the top of the liquid while the waste particles fall. This difference in effective density in water then enables the concentration of copper to take place. In this step, the first revenue generating material is produced (copper concentrate), which is typically what is sold by the miners (rather than the metal itself).

Pyrometallurgical reduction: Further treatment of copper concentrate needs to take place to achieve two ends — the reduction of the ore to metal ratio and the removal of further gangue material. In the first step, copper concentrate is mixed with various fluxes and fuels. The mix is then heated to a temperature of around 1,200°C, at which point a copper matte is separated from a silicate slag in which any residual gangue dissolves.

Conversion: The copper matte is blown with oxygen in a converter. This step realizes copper metal for the first time, with the sulphur in the matte being released as sulphur dioxide and the resulting blister copper achieving a purity of between 97% and 99% Cu.

Electrolytic refining: The blister copper is taken to a refinery where it will serve as the anode in an electrolytic refining process. The anode is immersed in a solution of sulphuric acid and copper sulphate. A current is then passed through the anode, causing it to dissolve into solution. The copper ions are then deposited

Underground Mines Carry Much Greater Ore Extraction Costs Than Open Pit Mines as the Latter Requires Less Energy and Are Scalable

There Are Two Routes of Copper Extraction: Milling and SxEw; Milling Accounts for ~80% of the Current Mine Supply



in a pure form (99.99% Cu) at the copper cathode, while any residual impurities from the anode are left behind.

The alternative route was first developed at the Bluebird mine in the U.S. in the 1980s. It was introduced as a way (initially) to exploit oxides that had hitherto been considered waste material in the process of mining sulphide ores to be treated through the milling route. Copper oxide materials arise as a result of the natural weathering of a sulphide outcrop. Alternatively, they can emerge as a consequence of the weathering achieved as low-grade waste copper sulphides are stockpiled and thus exposed to the elements. They can also arise under the action of bacterial agents such as Thiobacillus Ferrooxidans. Rather than simply being a mechanism to exploit waste material, SxEw copper production now accounts for ~20% of mined production. The main elements in this process are described below and shown schematically in Exhibit 46. Mining: As in the sulphide milling route, this process involves the separation of

ore containing sufficient quantities of valuable metal-bearing mineral from waste rock.

Crushing: Rather than requiring the physical liberation of different mineral types within a volume of ore, the SxEw route can proceed with coarser-sized material. Consequently, only initial crushing rather than grinding is required.

Acid leaching: The crushed oxide ore is placed on a leach pad and treated with a weak acid solution, into which copper then dissolves. The copper-rich solution, rather appealingly known as pregnant liquor, is then collected and sent to the solvent extraction stage of the process.

Solvent extraction: This step aims to increase the copper concentration in the solution to such a level that electrolysis and deposition of copper can be achieved. To this end, the pregnant liquor is first contacted with an organic solvent. Copper passes into the solvent, thus restoring the original acid that is subsequently recycled to the leach side. The organic solution is then itself stripped of its copper by reacting with a concentrated acid solution. This returns the organic reagent, which can once again be recycled. The concentrated copper/acid solution then has a copper concentration high enough to proceed to the final phase of production.

Electrowinning: It is equivalent to the electrolytic refining process described earlier, except that the copper is contained in solution rather than having to be introduced via the anode. A current is passed through the copper solution extracted previously and pure copper is deposited at the cathode.

As mentioned previously, copper oxide materials are often found as a weathered cap at the outcropping of a primary copper ore body. This weathering process is also responsible for another (arguably more important) feature of mined copper production — the secondary or supergene enrichment (see Exhibit 47). The action of water on an outcropping of sulphide material oxidizes and also leaches that material (through the process of naturally forming acid solutions, as the water reacts with sulphide minerals). This leaching results in copper dissolving out of the oxide layer and travelling down through the ore body, until it hits the water table. At this point, the copper in solution is re-precipitated out, resulting in the enrichment of the ore at the level of the water table. This enrichment creates a target for the miners. Specifically, targeting the secondary sulphide zone enables generation of higher cash flows early in a mine's life before moving on to the lower grade primary sulphide. This can have a very significant impact on the economics of a mine's development.



Exhibit 45 The Traditional Mining Route, Involving the Concentration of Sulphide Ores, Drives the Vast Majority of Mined Copper Production (~80%)

Source: Corporate reports. Exhibit 46 The SxEw Route Exploits Oxide Ores and Accounts for the Residual 20% of the

Mined Copper Production

Source: Corporate reports.



Exhibit 47 An Important Feature of the Sulphide Route Is the Ability to Take Advantage of High-

Grade Copper Ores in Secondary or Supergene Enrichment Zones; These Locations Can Provide Significant Additional Early Stage Cash Flows for a Miner


The Dependency of Global Copper Supply on Chile's Unique Geology Over the second half of the 20th century, the ability to supply the tremendous demand for copper at a low price depended on the exploitation of the massive copper porphyries in Chile. However, Chile's geology is unique and radically different from that of the second largest copper miner — China. In our view, the expectations that there will ever be a repeat of the Chilean copper boom are misplaced.

The geology of copper deposits is highly technical with a huge variety of physical and chemical processes at work in the development of the ore bodies that stand behind today's industrial production of copper. Very broadly, however, the deposits of economic interest may be grouped into three types. Porphyry deposits: These are deposits associated with volcanism in general and

with the tectonic subduction zones (one continental plate moving beneath another) in particular along the eastern edge of the Americas and around the entire perimeter of the Pacific Basin. These deposits are hydrothermal in nature and rely on the mineral solvency and concentrating abilities of pressurized high-temperature water. Hot aqueous solutions circulating through the Earth's crust dissolve the minerals contained in the host rocks through which they circulate. As these solutions cool down, the dissolved minerals are precipitated out of the solution. When the hydrothermal solution is rich in an economically interesting (i.e., scarce) metal, the precipitated concentration results in a mineable ore body. Porphyry deposits consist of numerous fractures (typically these fractures being millimeters in width separated by centimeters in the host deposit) resulting from a magmatic intrusion into a host rock. These fractures form the veins through which hydrothermal solutions were able to escape from the Earth's crust and, in doing so, undergo the process of cooling and mineral precipitation. The economic implication of the porphyry form is that selective mining (whereby individual

Primary Sulphide Ore

Secondary Enriched Sulphide Ore

Leached or Oxidised Zone

Water Table

Surface

Action of water in dissolving sulphide mineralisation

Supergene Zone

Hypogene Zone

Knowing Chile's Influence on Global Copper Supply Allows Us to Form Expectations About the Viability of Future Supply Growth Accordingly by Factoring the Absence of "Another Chile'' Emerging on the Horizon



high-grade veins are extracted) is impossible. Rather, bulk mining techniques must be employed. Consequently, this step marks the breakthrough from small-scale labor-intensive copper mining to the large-scale capital-intensive mining techniques employed today. As a piece of trivia, it was at Rio's Bingham Canyon that the first-ever demonstration of economic bulk copper mining was made. The existence of low-grade but massive copper porphyry deposits mined by bulk techniques stands behind the use of copper on the scale that we see in industrial societies today (see Exhibit 49). Without this step, the intensity of power and electricity use that fuels modern society would be impossible.

Massive sulphide deposits: These are a second class of hydrothermal deposits that rely on the enrichment properties of hydrothermal solutions for their ultimate origin. However, the mechanism by which the precipitation of minerals occurs is markedly different. Massive sulphide deposits are created in submarine environments where a volcanic phenomenon leads to the direct expulsion of a sulphide and metal-enriched hydrothermal solution into the ocean. The rapid cooling of the hot solution leads to mineral precipitation and the formation of a "blanket" of sulphide material around the vent. The portions of the ocean's crust where this process has taken place (and that are now above sea level) can be mined for copper and other minerals. The fact that the hydrothermal solution is expelled directly into the ocean explains why there is very little gangue material present in these deposits. The term "massive" is intended to reflect this mineral's concentration rather than the size of the deposit per se. This process of ore formation is ongoing today on the ocean's floor through the medium of "black smokers," whose name arises as a consequence of the sulphide precipitation having the appearance of soot. As another piece of trivia, the word copper ultimately derives from the Greek word (and island) Cyprus, where copper was mined in ancient times from a massive sulphide ore body.

Sediment-hosted deposits: The last broad class of ore bodies is sediment hosted copper or stratiform sedimentary deposits. Unlike the previous two types that have an intrusive nature and are associated with volcanic activity, sedimentary hosted deposits are found in marine sedimentary rocks and have a characteristic of being flat (or layered) in nature. They can extend horizontally over a very significant area. The most famous stratiform deposit — the Kupferscheifer of Northern Europe, which has been mined continuously since the 14th century — extends over 6,000 square kilometers but averages just 20 centimeters in depth! The origin of these deposits is controversial, but the most likely explanation is the leaking of hydrothermal fluids into sediments during the deposition either at or before the consolidation of the sediment into rock. The Zambian copper belt and DRC deposits fall within this classification.

For the sake of completeness, we also include a description of two further types of copper deposits. However, neither of them is as important economically as the three categories discussed earlier. Vein deposits: These deposits are tabular in nature with sharply defined

boundaries distinguishing valuable ore from the worthless host rock. They arise due to the presence of a clear fracture within a host rock through which hydrothermal fluid flows, with the fracture being filled with ore through the deposition of dissolved sulphides over time. These deposits can be incredibly rich but are (quantitatively speaking) very small. The deposits found in Cornwall, which played a pivotal role in the early industrial history of the U.K. but are now only of historical interest, belong to this category.

Magmatic segregation deposits: These deposits are unusual in not having anything to do with the circulation of hydrothermal solutions and their ability to concentrate copper-bearing minerals. During their formation, certain magmas, upon rising through the Earth's crust, become saturated with iron sulphide (FeS). As the magma begins to cool down, the iron sulphide forms into droplets that sink through the less dense host magmatic solution, forming a molten iron sulphide solution at the bottom of the magma chamber. As this solution solidifies, it forms



a mass of pyrrhotite (FeS) dotted with grains of chalcopyrite (CuFeS2) and also nickel-bearing minerals. The Sudbury basin is an example of this type of deposit, where copper exists as a fairly significant by-product of the nickel mining operation.

Returning to the three most important types of copper deposits, we show the grade to tonnage distribution of the 1,400 known copper deposits in Exhibit 48. The inverse relationship between grade and size is readily apparent. We see massive sulphide deposits fairly evenly distributed at the higher grades, sedimentary hosted deposits in the middle of the grade tonnage distribution, and copper porphyry deposits at the far end being the largest but the lowest grade of the three types (see Exhibit 49 through Exhibit 51). As is evident in Exhibit 51, the scale overwhelms quality as far as the availability of metal is concerned. 80% of the copper identified for possible exploitation sits within the lowest-grade ore bodies (and it should be remembered that costs are inversely related to grade). Given that the world's copper supply relies on the economic exploitation of porphyry deposits, it must be uniquely reliant on one country in particular. This country is, of course, Chile.

Exhibit 52 sets out the endowment of copper by country and by ore type. Exhibit 53 shows the total cumulative copper that this represents. The purpose of this analysis is to make the unique position of Chile clear. Not only is Chile the world's largest producer of mined copper metal (nearly 4,000ktpa more than the second largest producer — China), but it is uniquely well-endowed with the metal. At first instance, the industrial-scale consumption of copper on the back of mass electrification programs was permitted by the exploitation of the U.S. copper endowment (second only to Chile). Subsequently, electrification on a more global scale was occasioned by the development of the Chilean deposits post the market reforms of the 1970s and 1980s.

This analysis also highlights a few further features. First, as rich as the deposits of the DRC and Zambia are, they cannot possibly support the world's copper needs. The contained metal of these locations is simply too small. It is not only the political risk but also the geological endowment that will limit the contribution that these countries can make. While there will be significant value created for those that can enter these jurisdictions and take advantage of the incredibly high grade on offer (Glencore Xstrata within our coverage), this does not imply that Africa will ever be able to challenge the position currently occupied by Latin America. The second point is that Peru is not a new Chile. Peru is at best a "Chile lite." Despite the grandiose claims in some quarters as to what Peru may be capable of delivering, the ultimate output of any location depends on its geology. While Peru has significant room for expansion, this does not imply that it will ever be able to replicate the impact that the Chilean tons had on global copper prices.

There is also more to Chile than just the absolute magnitude of its total endowment. Specifically, it is the individual scale of the deposits within the country (see Exhibit 54). Chile is not only blessed with the lion's share of the world's copper, but also the copper that it has comes readily packaged in the most convenient possible form — massive ore bodies with high inherent mining optionality. Scale in mining creates options and the possibility of numerous exploitation and development patterns. All other things being equal, it makes mining easier. Once again, no other location will present as much in the way of a "low hanging fruit" as was on offer during the development of the Chilean copper industry. The future is going to be far harder than the history of the last few decades might suggest. Consequently, an analysis of the historical development of Chile's deposits will offer some important conclusions for how the future development of global copper supply may proceed.

Chile Is the World's Largest Copper Producer and Has the Best Endowment of the Metal; It Is Hard for DRC, Zambia and Peru to Ever Compare



Exhibit 48 In the Distribution of Known Copper Deposits, the Graph Shows the Typical Inverse Relationship Between Grade and Size, With Copper Porphyry Occupying a Place of Privilege at One End of That Distribution


0.03

0.06

0.13

0.25

0.50

1.00

2.00

4.00

8.00

16.00

0.01 0.1 1 10 100 1000 10000 100000

Ore

Gra

de

(%

Cu

)

Ore Tonnage (Mt)

Distribution of the World's Main Copper Deposits

Massive Sulphide Sediment Hosted Porphyry



Exhibit 49 Copper Porphyries Represent the Lowest Grade But the Most Abundant Source of Copper Supply; Their Exploitation by Bulk Mining Methods Enabled the Development of the Modern Power-Intensive Industrial Society

Source: USGS and Bernstein estimates and analysis. Exhibit 50 The Average In Situ Geological Abundance in Copper Porphyry Deposits Is 0.5%,

Which Has Important Implications for the Long-Term Grade Profile of Copper Production; Anything Higher Than This Must Be Temporary


0

50,000

100,000

150,000

200,000

250,000

300,000

350,000


Mt O

re

Ore Tonnage by Deposit Type

0.0%

0.5%

1.0%

1.5%

2.0%

2.5%


Co

pp

er

Gra

de

Average Ore Grade by Deposit Type



Exhibit 51 The World's Consumption of Copper Is Predicated Upon the Exploitation of the Lowest-Grade Copper Deposits


Exhibit 52 Poland, Zambia and the DRC Are High Grade But Too Small to Displace American Preeminence


0

200

400

600

800

1,000

1,200

1,400

1,600

1,800


Co

nta

ined

Me

tal (

Mt)

Contained Copper by Deposit Type

2.2

2.4

2.6

2.8

3.0

3.2

3.4

0.6

0.8

1,4001,3001,200400 500

1.6

0.2

0.0

2,100900 2,0001,9001,8001,7001,6001,500600 1,1001,0008007003000 100 200

1.4

1.2

1.0

1.8

0.4

2.0

Sediment Hosted

Massive Sulphide

Porphyry

Ore

Gra

de

-%

Cu

Contained Metal - Mt

DRC

Zambia

Poland

ChileUnited States

Peru China



Exhibit 53 Chile's Dominance in the Ability to Supply the World's Copper Demand Is Clear

Source: USGS and Bernstein estimates and analysis. Exhibit 54 Chile Is Not Only Uniquely Well-Endowed in Absolute Tons, But the Size of Its

Deposits Is Unmatched


900 1,2001,1001,000

150

800700600500400300

200

250

300

350

400

450

500

550

600

650

700

2001000

50

100

0

2,1002,0001,9001,8001,7001,6001,5001,4001,300

Sediment Hosted

Massive Sulphide

Porphyry

Contained Metal ‐Mt

Global Cumulative Contained Metal ‐ Mt

Global Copper Endowment

Chile

USA

Peru

DRC Zambia

Canada

Poland

China

0

5

10

15

20

25

Co

nta

ined

Cu

(Mt)

Average Copper Porphyry Size



Low Copper Prices During the 1980s and 1990s Were Predicated Upon Chile Between 1983 and 2003, the average real copper price was just US$3,250/t. Meanwhile, Chile increased copper production by 3,700ktpa. Since 2003, the largest source of copper growth has been China — a country with an endowment just 8% that of Chile's!

Chile's contribution to the copper supply is hard to exaggerate — it is 260% larger than its nearest rival China (see Exhibit 55). As the previous analysis has made clear, it is Chile's unsurpassed geological endowment that makes this possible. However, in terms of supply growth, the last decade has seen Chile stall and China emerge as the fastest-growing producer (see Exhibit 56). Exhibit 57 through Exhibit 66 show the history of the last century of supply from today's 10 most important locations. These exhibits highlight that in five of the top 10 producing countries, mined growth has stalled and in some cases even declined. Moreover, the countries wherein growth has stalled (Chile, the U.S., Australia, Canada and Commonwealth of Independent States [CIS]) represent the "easy" political locations for mining investment. That is, the mining industry invested in, and developed, the most attractive deposits from risk-return perspective geology first (which makes perfect sense). This serves as yet another reminder of the difficulty that future supply growth will encounter, as the copper industry is forced to take on ever higher risk to secure new sources of supply. These exhibits also help contextualize the recent increases we have seen in output from Africa. In the DRC and Zambia, copper production growth has been strong but only in so far as it corresponds to the recovery from the catastrophic political disruptions — the Great Lakes conflict in the DRC and the nationalization of the Copper Belt in Zambia. Production today has only just surpassed the levels reached prior to these political events.

In the Past, the Mining Industry Invested in and Developed the Most Attractive Deposits, in Risk-Return Terms; This Serves as a Reminder of the Difficulty That Future Mined Growth Will Encounter

Exhibit 55 Chile Stands Out as the Most Important Source of Copper Supply; However, This Was Not Always the Case, and the Study of Chile's Development Is Critical to an Understanding of the Future Copper Price Trajectory

Source: Wood Mackenzie and Bernstein estimates and analysis.

0

1000

2000

3000

4000

5000

6000

Chile China Peru USA Australia Zambia (& Northern

Rhodesia)

Russia Congo DR (Zaire & Bel

Congo)

Canada Mexico

Min

ed C

op

per

Pro

du

ctio

n (k

t)




Exhibit 56 While Chile's Growth Has Slowed Down Over the Last Decade, China's Has Accelerated


Exhibit 57 It Was the 20 Years from 1980 to 2000 That Saw Chilean Copper Growth Explode

Exhibit 58 However, Since 2000, Chile Has Stagnated and China Emerged as the Second Largest Producer of Mined Copper

Source: Wood Mackenzie, Mitchell and Bernstein estimates and analysis.


-4%

-2%

0%

2%

4%

6%

8%

10%

China Peru Chile Australia Mexico DRC Zambia Russia Canada USA

CA

GR

199

0-20

03

Mined Output Growth of Largest Copper Producers

-

1,000

2,000

3,000

4,000

5,000

6,000

1900

1905

1910

1915

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

kt C

u

Chile Mined Copper Output

-

200

400

600

800

1,000

1,200

1,400

1,600

1,80019

0019

0519

1019

1519

2019

2519

3019

3519

4019

4519

5019

5519

6019

6519

7019

7519

8019

8519

9019

9520

0020

0520

10

kt C

u

China Mined Copper Output



Exhibit 59 Peru's Growth Has Been Sporadic Exhibit 60 The U.S. Has Played a Critical Role in the Copper Industry and Was, for a Long While, the World's Largest Producer



Exhibit 61 As With Chile, Australian Output Has Reached a Plateau

Exhibit 62 Meanwhile, Zambia Has Only Just Recovered from the Effects of the Previous Nationalization



-

200

400

600

800

1,000

1,200

1,40019

0019

0519

1019

1519

2019

2519

3019

3519

4019

4519

5019

5519

6019

6519

7019

7519

8019

8519

9019

9520

0020

0520

10

kt C

u

Peru Mined Copper Output

-

500

1,000

1,500

2,000

2,500

1900

1905

1910

1915

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

kt C

u

USA Mined Copper Output

-

100

200

300

400

500

600

700

800

900

1,000

1900

1905

1910

1915

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

kt C

u

Australia Mined Copper Output

-

100

200

300

400

500

600

700

800

1900

1905

1910

1915

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

kt C

u

Zambia Mined Copper Output



We also chart the history of global copper mining somewhat differently from the simple chronology of the last few exhibits, i.e., as the tale of the U.S. and Chile alone. For the first 80 years of the 20th century, the U.S. was the world's largest producer of copper. In fact, it was leading by a greater degree than that seen in Chile today. In 1900, the U.S. produced over 400% more copper than its nearest rival Spain (see Exhibit 67). The subsequent history of copper can be thought of as the transition of the title of the world's copper hegemon from the U.S. to Chile (see Exhibit 68 through Exhibit 72).

Exhibit 63 The Same Stagnation That We See in Chile (Once the Limits of Geology Are Reached) Is Evident in the CIS

Exhibit 64 The Impact of the Great Lakes Conflict on DRC Output Is Painfully Clear



Exhibit 65 Canada, While Still a Significant Producer, Is in Decline

Exhibit 66 Mexico Is Likely to Become a More Important Producer Over Time



The Unique Features of the Last Century Have Been the Transition of Copper Supply from the Second- to First-Most Endowed Country Accompanied by a Change in Bulk Mining Technology

-

200

400

600

800

1,000

1,200

1,40019

0019

0519

1019

1519

2019

2519

3019

3519

4019

4519

5019

5519

6019

6519

7019

7519

8019

8519

9019

9520

0020

0520

10

kt C

u

CIS Mined Copper Output

-

100

200

300

400

500

600

700

1900

1905

1910

1915

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

kt C

u

DRC Mined Copper Output

-

100

200

300

400

500

600

700

800

900

1900

1905

1910

1915

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

kt C

u

Canada Mined Copper Output

-

50

100

150

200

250

300

350

400

450

500

1900

1905

1910

1915

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

kt C

u

Mexico Mined Copper Output



As a result, the last century involved a movement of copper production from the geology of the world's second-most endowed country — the U.S. — to the world's most endowed country — Chile. In addition, the mining technology behind copper production has advanced substantially, and the change is intimately tied up with the nature of copper porphyry deposits. In Exhibit 73, we show one of the first uses of industrial capital equipment on a copper mine in Australia at the start of the 19th century. Prior to this, copper mining far more frequently took place underground, with human labor chasing rich streams — an activity that is inherently less productive and economic only insofar as labor is abundant and cheap. In Exhibit 74, we show the modern day equivalent of the early steam shovel, capable of moving 110 tons of material in each movement. The most striking feature of the modern vehicle versus its predecessor (apart from the improvement in color scheme) is, of course, size. A vast increase in scale and accompanying mechanical efficiency has been achieved over a century or so from the introduction of mechanical shovels into mining. However, as important as the increase in scale is, the basic concept has remained unchanged. Overall, the radical and discontinuous transition in mining took place when capital displaced labor and bulk mining displaced selective mining. Since then, mining has witnessed marginal improvement of the same basic underlying idea.

We summarize both of these trends in Exhibit 75, which looks at the history of copper porphyry discoveries, and the impact of the discoveries in the U.S. and, subsequently, in Chile on the real copper price. The real price of copper halved in a decade on the back of an explosion in copper porphyry exploitation at the start of the 20th century. The demonstration of the economic viability of low-grade copper mining encouraged the delineation and discovery of a significant number of very large deposits that had hitherto been thought un-mineable using human muscle power alone. In the post-war period, as the scale of new copper porphyries began to decline, the real price of copper began to rise. Against a rising demand environment (supported by electrification programs in the West), the technological and geological step-change of new porphyry deposit discoveries began to run its course.

Exhibit 67 In 1900, the U.S. Played a Very Similar Role in the Global Copper Supply as Chile

Does Today; the Development of the Supply Side of the Copper Industry Over the Last Century Is the History of the Volume Transition from the U.S. to Chile


0

50

100

150

200

250

300

USA Spain Japan Chile Australia Mexico Germany Canada Peru CIS

Min

ed C

op

pe

r (k

t)

1900 — Copper Supply by Origin



Exhibit 68 By 1925, the Importance of Chile Was Starting to Become Clear...

Source: Wood Mackenzie, Mitchell and Bernstein estimates and analysis. Exhibit 69 ...Though Africa (Zambia and DRC) Have Always Had a Role to Play in the Supply of

Copper


0

100

200

300

400

500

600

700

800

USA Chile Congo DR (Zaire & Bel

Congo)

Japan Mexico Canada Spain Peru Germany Cuba

Min

ed

Co

pp

er (

kt)


0

100

200

300

400

500

600

700

800

900

USA Chile Zambia (& Northern

Rhodesia)

Canada CIS Congo DR (Zaire & Bel

Congo)

Mexico Serbia (& Yugoslavia)

Japan South Africa

Min

ed

Co

pp

er

(kt)




Exhibit 70 By 1975, the Dominance of the U.S. in the Supply of Copper Was Already Challenged

Source: Wood Mackenzie, Mitchell and Bernstein estimates and analysis. Exhibit 71 However, It Was the Market Reforms That Inaugurated the "Miracle of Chile" That

Saw the U.S. Finally Topple as the Superior Geology and Lower Labor Costs Established Chile as the Leading Global Copper Producer


0

200

400

600

800

1,000

1,200

1,400

USA CIS Chile Canada Zambia (& Northern

Rhodesia)

Congo DR (Zaire & Bel

Congo)

Poland Philippines Australia Peru

Min

ed

Co

pp

er (

kt)


0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

5,000

Chile USA Indonesia Australia Canada China Russia Peru Poland Kazakhstan

Min

ed

Co

pp

er

(kt)




Exhibit 72 Despite All of These Changes, Copper Production Remains Highly Concentrated in Very Few Regions

Source: Wood Mackenzie, Mitchell and Bernstein estimates and analysis. Exhibit 73 An Early Steam Shovel at Mt. Morgan Copper Mine in Australia at the Turn of the 20th

Century

Source: Wikimedia Commons.

70.0%

75.0%

80.0%

85.0%

90.0%

95.0%

100.0%

1900 1925 1950 1975 2000 2012

Top 10 Producers' Share of Global Copper Supply



Exhibit 74 An Electric Rope Shovel at the Turn of the 21st Century — Same Idea, Slightly Bigger Scale

Source: Rio Tinto. Exhibit 75 The Long-Term History of Copper Porphyry Discoveries Shows the Marked Impact

That These Deposits Have Had on the Real Copper Price; Structurally Falling Prices Have Been Associated With Increased Finds of Relatively Few Massive Ore Bodies; We Are Not Currently in Such a Situation — There Is No New Chile on the Horizon

Note: Bars are the average size of new copper porphyry discoveries and the line is the copper price.


3,000

4,000

5,000

6,000

7,000

8,000

9,000

10,000

0

2

4

6

8

10

12

14

16

18

Co

pp

er P

rice

-U

S$/

t -R

eal 2

011U

S$

Ave

rag

e S

ize

of N

ew C

u P

orp

hyr

y D

isco

very

-M

t Cu



An observer in the 1970s would have been forgiven for thinking that real copper price increases were likely to continue indefinitely. However, two things intervened to change this. First, the Western copper demand underwent a step-change post the oil shocks. Second, the supply side saw the emergence of Chile. Chile's economic reforms of the late 1970s and 1980s drove the return of foreign investment and saw a new wave of copper deposit discoveries. Escondida (discovered 1981) and Collahuasi Rosario (1985) are just a few deposits discovered then. The supply of new material from Chile was sufficient to instigate two decades of negative copper price performance. However, since that time, the rate of new large copper porphyry discoveries has ground to a halt.

Fast forward to today. Demand for commodities has accelerated on the back of China's industrialization, and we are back in the territory of real price increases. Or rather, we are back to where the trend line of price increases would have been if extrapolated out from the end of the 1980s and the impact of Chile was removed. Clearly, this high-level argument is insufficient to justify a price forecast, but it does help set the scene for the arguments to come.

As we have discussed, there is no new Chile on the horizon. Peru is the closest comparison, and even its geological endowment is less than half that of Chile. While Zambia and the DRC undoubtedly have high-grade deposits, they are simply not large enough to play the role of Chile in the global copper supply.

Considered over a very long term, there have been two periods of significant real-term price declines in copper driven by two discrete supply side events (ignoring demand-side events for the time being). Introduction of bulk mining techniques on the copper porphyry's of the U.S. at the

start of the 20th century. The exploitation of Chile's superior geology using these techniques during the

1980s and 1990s. These two factors first established the U.S. as the world's leading copper

producer and subsequently displaced it in favor of Chile. In both cases, the transition was achieved through the supply of significant new low-cost volumes of metal and was mirrored by a period of sustained low copper prices. However, this twofold transition was also one that established Chile's rightful place (given its geology) as the world's premier copper producer. The history of copper over the last century is the history of production from the world's two most well-endowed supply locations. All subsequent history will come from regions with inferior geology and more challenging political and technical environments. Consequently, we struggle to see how new supply can be unlocked with stagnant or declining copper prices.

China Versus Chile A very strong relationship exists between a country's geological endowment and its production of copper. It suggests that China is producing far too much metal and that there is very limited additional upside available from Chile.

To understand just how difficult the future of copper mining will be outside of the U.S. and Chile, we perform another piece of analysis looking at copper endowment and production (see Exhibit 76). Understandably, there is a very strong relationship between the percentage of the world's output that is attributable to a particular location and the percentage of the world's copper that is present there. A country tends to produce more metal to the extent that the metal is in the ground waiting to be developed. However, it is the departures from this relationship that are interesting. They tell us which countries are producing too much metal relative to their geological endowment (in which case the production must be under threat if costs in that location begin to rise) and which ones have potential headroom for further expansion (see Exhibit 77).

The exhibits highlight the anomalous position occupied by China, whose copper production is significantly out of proportion with its underlying geology. We believe that this situation has arisen only as a consequence of the recently high

Previous Increases in Real Copper Prices Were Reversed in the 1980s on the Back of a Step-Change in the Western Copper Demand and Chile's Economic Reforms; Introduction of Bulk Mining Techniques in Copper Porphyry Mining Aided the Process Considerably

China Is Producing Far Too Much Metal Given Its Geological Endowment, While Chile Has Reached a Plateau



copper prices and the low-wage environment in China relative to other mining jurisdictions. If this interpretation is correct, it highlights the fragility of the world's current mine supply to falling prices. The majority of growth in mined supply observed over the last decade has come from a country with a very limited endowment of the metal. In addition, this country will face increasing mining costs as the returns to labor (and away from capital) begin to take effect and the Lewis tipping point is reached, at which the rate of capital formation begins to slow down.

However, so far the discussion of Chile has focused on a description rather than an explanation of the growth that the country has enjoyed. In order to really understand the implications of the Chilean story for other regions, we must now explain the growth.

Exhibit 76 If We Chart Countries' Current Copper Supply vs. Their Underlying Endowment, We

See an Understandably Strong Relationship; Good Geology Tends to Imply Easy Mining

Source: USGS, Wood Mackenzie and Bernstein estimates and analysis.

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

35.0%

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0%

% of P

rodu

ction

% of Identified Copper

Chile

USA

China



Exhibit 77 It Is the Departures from This Relationship That Are Interesting; Chile Is No Longer an Easy Win and the Largest Source of Mined Copper Growth; Over the Last Decade, China Has Been Producing at More Than Twice the Level Its Geology Would Suggest

Source: USGS, Wood Mackenzie and Bernstein estimates and analysis.

Future Copper Development Will Prove Harder Than Many Anticipate Chile was able to increase the production of copper through two mechanisms: more rapid exploitation of deposits already known and new deposit discoveries. The application of the first process to the locations outside of Chile yields only one-third the required copper while the second process is simply not working.

There are two basic mechanisms by which a country can increase its output of any mined commodity: Discovery — In the first place, a country can discover more of the commodity in

question and increase its geological endowment as well as relative production at the same time. The classic example of this is Escondida — the world's largest copper mine discovered in 1981 and commissioned in 1990. It increased both the reserve base and production of Chile.

Development — As opposed to discovering more of a commodity, a country can choose to accelerate the development of the resources that it has. In this case, production increases beyond the level implied in the geological endowment of the country. This is what has happened in China.

If we go back to the very beginning of the Chilean copper industry, we can see both of these effects in operation. In Exhibit 78 and Exhibit 79, we see the enormous increase in exploitable material occasioned by the ability to target the lower grades of material contained in copper porphyry deposits through the application of capital rather than labor. As with copper porphyries in general, the increase in ore tonnage more than offsets the decreases in grade and the contained metal increases sharply (see Exhibit 80). The increase in production out of Chile has been greater than the increase in contained metal (see Exhibit 81). Between 1935 and 2012, Chile's exploitable metal increased 630% while production increased 1,970%. The increase in available metal clearly indicates the significant role the exploration and the discovery of new deposits have had on Chile's output. However, the greater increase in output tells us that accelerated exploitation of existing reserves also contributed. This is seen most clearly in the reserve life of

0.0

0.5

1.0

1.5

2.0

2.5

3.0

China Indonesia Russia Peru Mexico Zambia Poland Chile Canada DRC USA

>1

Imp

lies

Pro

du

cti

on

in E

xces

s o

f En

do

wm

en

t

Copper Production to Copper Endowment of 10 Largest Producers

More Rapid Exploitation of Existing Deposits Would Fail to Clear Demand While New Discoveries Have Come to a Complete Halt; Hence, an Increase in Real Copper Price Seems to Be Unavoidable



Chile, which was sufficient to support nearly 100 years of output in 1935 and has fallen to just over 30 years today (see Exhibit 82).

The fundamental reason for the acceleration in mined output above the rate of new discoveries, with the corresponding reduction in mine life, is that it makes economic sense. In any discounted cash flow model of mine value, the far out years are so highly discounted that they add very little value today. From a value perspective, there is no point in having 100 years worth of mine supply above, say, 50 years. However, if those extra years can be brought forward so that they count towards today's production, then very significant value is unlocked. Investing capital to double the rate of exploitation (say through increasing milling capacity) while reducing the mine life enables shareholders to benefit today from tomorrow's production. This is clearly what has happened in Chile. However, this process cannot continue indefinitely. Once there is roughly 30-year life of mine left, all the years of production are relevant to the value proposition of the mine. It is no longer the case that some years are so far in the future so as to be essentially worthless. Consequently, the expenditure of capital in an attempt to bring those years forward destroys rather than creates value. Clearly, the amount of value creation or destruction depends on the intensity of capital that must be expended to accelerate production. Exhibit 83 shows how this trade-off works. The lower the capital intensity, the easier it is to create value through accelerating production. The critical point is that when the reserve life of a country reaches between 25 and 35 years, there is no value to be gained from increasing the rate of exploitation of existing deposits. Exhibit 84 and Exhibit 85 show how the investment case for doubling capacity and halving life changes depending on the original mine life. The critical point for Chile is that the country has passed this threshold — the rate of growth in its resource base has slowed, the rate of growth in production has increased, and Chile's current life of reserves suggests that it will struggle to keep track with depletion, let alone grow through more efficient exploitation of existing reserves. This mechanism stands behind the stagnation of Chilean mined output at ~5,500ktpa. It also provides the fundamental explanation why a country's output of a commodity should be given by its underlying geological endowment. Past this critical point, all subsequent increases in mined growth must come from one of the two sources. The exploitation of resources that did not originally pass the economic filter to be

included in reserves (thus it requires much higher prices to render the resources into economically viable sources of production).

New discoveries that enable the resource base to increase in proportion to the increase in mined production.

Having understood the financial mechanism that generates the coupling between production and reserves through an analysis of Chile's copper mining history, we are in a position to extend it globally. Exhibit 86 shows the current reserve life of the 10 largest copper producers (which collectively account for more than 80% of supply). As can be seen, there are some locations — notably China — where known reserves are woefully short of current production, and others — such as Peru — where there is significant upside. This enables us to calculate the increase in mined production that the politically unimpeded development of a country's geology should allow. We then predict the ultimate trajectories for peak production for those countries whose history we have shown in Exhibit 57 through Exhibit 66. We look at this for every country with known copper deposits. The outcome of our analysis for the top 10 most significant new supply locations is shown in Exhibit 87. It indicates that current copper reserves have the ability to supply less than one-third of the world's incremental copper demand by 2030.

Frontloading Cash Generation Unlocks Significant Value Potential; However, If We Frontload Cash at the Expense of Reducing LOM Below 25-30 Years, the Exercise Becomes Value Destructive as Opposed to Value Generating



In sum, either more production will have to come from material currently defined as resources rather than reserves, which will necessitate a price higher than today's, or new supply will have to come from fresh discoveries. However, if we look at the recent history of new copper porphyry discoveries, two things become clear. First, the rate of discovery of massive new ore bodies has declined sharply. Second, the lead time to development of the deposits that are actually capable of making a difference to the supply/demand dynamic has increased dramatically (see Exhibit 88). It took 10 years to bring Escondida on line. It will have taken 17 to develop Oyu Tolgoi (and probably more like 20-25 years for it to realize its full potential). Finally, Pebble is at 25 years and counting. Consequently, new discoveries are incapable of meeting the world's copper demand. That leaves one alternative — namely, higher prices and the ability to supply the world from deposits that do not meet investment thresholds at today's prices.

Exhibit 78 The History of Chilean Copper

Development Is One of Massive Increases in Resources...

Exhibit 79 ...Occasioned by the Ability to Exploit Ever Lower-Grade Material

Source: Wood Mackenzie, ABMS and Bernstein estimates and analysis.


Exhibit 80 However, the Net Result Is a Massive Increase in Available Metal...

Exhibit 81 ...And an Even Greater Rise in Metal Output; Hence, There Has to Be More to Chile Than Increasing Discovery Rates...



0

5,000

10,000

15,000

20,000

25,000

30,000

35,000

1935 2012

Ore

(M

t)

Chile Ore Reserves

0.0%

0.5%

1.0%

1.5%

2.0%

2.5%

1935 2012

Ore

Gra

de

(%)

Chile Reserve Grade

0

50,000

100,000

150,000

200,000

250,000

1935 2012

Co

pp

er in

Ore

(kt)

Chile Contained Copper in Reserves

0

1,000

2,000

3,000

4,000

5,000

6,000

1935 2012

Co

pp

er p

rod

uct

ion

(kt)

Chile Copper Production

Given That New Discovery Rates Have Declined Sharply and That Lead Time to Deposit Development Has Increased Dramatically, an Organic Increase in Copper Supply Will Be Possible Only When Copper Price Rises Sufficiently to Render Mining of Current Resources Economic



Exhibit 82 ...And There Is — It Is the Increased Efficiency in Exploiting Existing Material as Seen in Mine Life Reductions

Source: Wood Mackenzie, ABMS and Bernstein estimates and analysis. Exhibit 83 A Highly Non-Linear Relationship Exists in the Value Proposition Represented by

Mine Life Reductions; They Represent Efficiency Gains Down Only to ~30 Years LOM; Afterwards, They Become Value Destructive


0

20

40

60

80

100

120

1935 2012

Yea

rs

Chile Reserve Life

0%

20%

40%

60%

80%

100%

120%

140%

160%

100 90 80 70 60 50 40 30 20 10

Va

lue

Acc

reti

on

>1

00%

Im

pli

es V

alu

e G

ain

Reserve Life of Original Mine

Value Generation Through Doubling Mine Capacity and Halving Life of Mine

US$15,000/t Capital Intensity US$10,000/t Capital Intensity



Exhibit 84 For a Given Geology, Halving a Mine's Life from 50 Years to 25 Years Is a Highly Profitable Exercise

Source: Bernstein estimates and analysis. Exhibit 85 Halving a Mine's Life from 30 Years to 15 Years (from the Same Geology) Is a Value

Destructive Exercise


1500

2000

2500

3000

3500

4000

4500

5000

NPV Years 0-25 NPV Years 26-50 NPV Unexpanded Mine

NPV Gain on Bringing Years 26-50

Forward

NPV Loss Due to Capex

NPV Expanded Mine

NP

V (U

S$m

)Value of Expansion for 50-Year Life of Mine

1500

2000

2500

3000

3500

4000

4500

5000

NPV Years 0-15 NPV Years 16-30 NPV Unexpanded Mine

NPV Gain on Bringing Years 16-30

Forward

NPV Loss Due to Capex

NPV Expanded Mine

NP

V (U

S$m

)

Value of Expansion for 30-Year Life of Mine



Exhibit 86 This Goes Some Way to Explaining Why Chilean Metal Output Stalled After Hitting 5.5Mtpa and a 35-Year Average LOM; It Also Highlights Why Chinese Production Looks Challenged and Where the "Low(ish) Hanging Fruit" Lie

Source: Wood Mackenzie, ABMS and Bernstein estimates and analysis. Exhibit 87 Mine Life Expansions from Existing Reserves Have the Potential to Deliver Less

Than One-Third of the Required Copper Demand by 2030; Projects Exploiting New Resources Will Be Needed and This Requires New Finds of Copper


0

10

20

30

40

50

60

70

80

90

100

Chile China Peru USA Australia Zambia Russia DRC Canada Mexico

Ye

ars

Reserve Life of 10 Largest Copper Producers

0

200

400

600

800

1,000

1,200

1,400

1,600

Peru Mexico USA Mongolia Indonesia Pakistan Brazil Canada Iran Poland

Ad

dit

ion

al C

op

per

(kt

)

Additional Production Potential in Moving to 35-Year Reserve Life



Exhibit 88 However, the Exploitation Is Looking Ever Less Likely to Yield Results; Even If Massive New Finds Are Encountered, the History of Pebble and Oyu Tolgoi Tells Us That It Will Take 20-30 Years for These Finds to Deliver Commercially Meaningful Metal (Compared to 10 Years for Escondida)


The Structure of New Supply — It Will Struggle at Current Price Levels We examine the cash and capital costs of approximately 300 greenfield projects and the associated brownfield expansions. Unsurprisingly, there is significant operating margin to be made through capacity expansion. However, we also note that the capital requirement is such that, against a fully loaded discount rate (i.e., including an adjustment for country risk), it is only at prices above US$8,000/t to US$9,000/t that the majority of them will be value accretive (see Exhibit 91). Now, more than ever, shareholders are aware of the risks inherent in new greenfield projects (e.g., Pascua Lama, Minas Rio and Riversdale). Consequently, many investors have been demanding that capital is returned rather than expended in an effort to push commodity prices down. While we have not factored these strategic issues into our analysis of the returns required for new investment (basing them on our understanding of the investment protocols in the large mining houses), they support our belief that new copper project approvals will struggle. Moreover, in a period of declining or flattening commodity prices, concerns of value rather than growth dominate the thoughts of the miners. This acts as a break on new volume growth.

As with iron ore, the majority of the potential copper volume expansion belongs to a handful of projects (see Exhibit 92). However, the number of projects accounting for 50% of new copper capacity is larger for copper than for iron ore (i.e., 49 for copper versus 16 for iron ore). This raises an interesting side point — asset differentiation creates variation in costs between operations, thus leading to an inflection in the cost curve and hence margin creation. In other words, an asset is valuable to the extent that it is unique or at least enjoys a privileged position relative to other assets. If all assets in a commodity class were exactly the same, the industry would be unattractive. The asset or project size relative to all assets or projects can be used as a proxy for asset attractiveness (at the very least indicating how amenable it is to the deployment of ever more massive bulk mining techniques and avoidance of the risk of selective mining). This enables us to make a relatively objective comparison between commodity classes and an assessment of their

0

10

20

30

40

50

60

70

80

90

100

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

Co

nta

ined

Cu

-M

t

History of New Copper Discoveries

Chile Canada China Ecuador Peru Kazakhstan Indonesia United States Australia Papua New Guinea Philippines Pakistan Mongolia

Escondida

Pebble

Oyu Tolgoi

It Is Only at Prices Above US$8,000/t to US$9,000/t That the Majority of Projects Will Be Value Accretive; Given the Current Price Environment, Miners Are Reluctant to Approve New Projects; We See This Situation Persisting Until the Copper Price Starts Rising



fundamental attractiveness (see Exhibit 89). The most differentiated commodities are iron ore and copper, followed by nickel and aluminum. This conclusion comes in striking agreement with the observed pricing behavior of these commodities (see Exhibit 90). We would, however, stress that this analysis is included more as an interesting observation rather than a fundamental derivation of why certain commodities outperform others. Nevertheless, it helps to make the case that geological differentiation is a key component of return generation in mining.

Exhibit 89 We Look at Project-Size Differentiation, i.e., the Difference Between Percentage of Mines and Supply

Exhibit 90 This Measure Shows a Striking Correlation to Commodity Returns

Source: AME, CRU, Wood Mackenzie and Bernstein analysis. Source: Bernstein analysis.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0% 10%

20%

30%

40%

50%

60%

70%

80%

90%

100

%

% o

f M

ine

s

% of Supply

Asset "Differentiation" by Commodity

Fe Cu Ni Al

Fe

Cu

Ni

Al

0%

5%

10%

15%

20%

25%

0% 10% 20% 30% 40%

An

nu

alis

ed A

ver

ag

e C

om

mo

dit

y R

etr

un

% of Projects accounting for 50% of Possible New Supply

Geological Differentiation vs. Annual Returns

Exhibit 91 Against a Fully Loaded Cost of Capital, Many of Today's New Projects Will Struggle to Create Value

Exhibit 92 The Majority of the World's Largest Projects Are in the Hands of Relatively Few Major Mining Houses


20,000

17,000

16,000

15,000

14,000

13,000

12,000

11,000

10,000

9,000

8,000

7,000

6,000

5,000

4,000

3,000

2,000

1,000

10,000 15,000 25,0000

0

5,000

Global Cumulative New Production – kt Cu

Incentive Price –US$/t

15,000 20,000 25,000

500

550

600

650

700

10,000

0

50

100

150

200

250

300

350

400

450

0 5,000


New Production –ktCu



While the incentive price curve is critical in analyzing forward-looking prices, we believe its role is often misinterpreted. It is true that an incentive to supply new mined volume has to be estimated. However, it is far from clear that the market delivers this via a guarantee return on capital. Rather, we believe that it is a perception of value creation engendered by current prices rising above the incentive price for new capacity that triggers the investment decision. Once this happens, it typically results in oversupply, leading to price deterioration and the emergence of the cycle. Thus, an incentive is delivered over the cycle through massive initial cash generation followed by long periods of low margin and effective stagnation. The only difference in today's commodity cycle versus the "normal" cycle is the degree of oversupply necessary to displace the Chinese domestic mining industry.

In the case of iron ore, this requires a significant volume of material. In the case of copper, the quantum is much smaller. However, we believe that there are a number of factors that differentiate copper from iron ore to the benefit of copper. In our view, even where capital discipline is lacking, geological barriers to entry provide greater protection against the threat of imminent oversupply. The world's largest iron ore supply location — Australia — has been able to grow output, while the world's largest copper supply location — Chile — has not been able to do the same (see Exhibit 93). We would highlight two factors behind this, both of which speak to a decline in geological copper quality. Head grade: The head grade of milled copper has fallen over 25% in the last six

years. This does both: increases mine operating costs directly via a reduction of actual metal content, and reduces concentrate qualities and mill productivity (see Exhibit 94).

Stripping ratios: Mines are getting older and deeper. In addition to the grade of the accessed material deteriorating, the process by which this material is obtained is getting more difficult. This puts upward pressure on operating costs for the industry as a whole and acts to support price.

That said, we would reiterate the view that no demand-side scenario and no degree of geological attractiveness is sufficient to protect an industry from the prospect of oversupply indefinitely if capital discipline is lacking in the investment decisions of the major incumbents. Given that the most significant projects belong to industry incumbents, for copper as for iron ore, the fate of value creation in mining is largely in the hands of the miners themselves. They can either create or destroy the conditions for the continuation of the "super-cycle," depending almost entirely upon their own understanding of the role they play in determining commodity prices.

Even if Capital Discipline for Copper Is Lacking, Geological Barriers to Entry Provide Greater Protection Against the Threat of Imminent Oversupply Relative to Other Commodities; Nevertheless, Capital Discipline Still Plays a Huge Role and So the Fate of Value Creation in Copper Mining Is Largely in the Hands of the Miners Themselves



Exhibit 93 Unlike With Iron Ore, There Are Some Genuine Supply Side Constraints That Have Prevented Chilean Copper Supply from Responding to Price Increases in the Same Way as Australian Iron Ore Did

Source: AME, CRU, Wood Mackenzie and Bernstein analysis.

Exhibit 94 In Mining, Grade Is King; It Acts as Geological Gearing of the Labor Cost; Halve the Grade and You Will Double (at Least) the Unit Cost of Metal Output


Fundamentally, we do not believe that there is any incremental value in industry consolidation unless industry incumbents act with awareness of the role they play in determining commodity prices. In other words, they have to explicitly understand that value destruction through price reduction has a far more significant effect on their value than value creation through volume growth.

We see three levers arising from this, through which the miners can control their impact on the markets and maximize their value.

0

20

40

60

80

100

120

140

160

180

200

2007 2008 2009 2010 2011 2012

Ind

exed

pro

du

cti

on

200

7 =

10

0

Chilean Copper vs. Australian Iron Ore

Chile - Copper Australia - Iron Ore

0.80

0.90

1.00

1.10

1.20

1.30

1.40

1.50

1.60

1.70

1.80

% C

u

Mined Copper Head Grade

Value Maximization Can Be Achieved Via More Explicit Awareness of the Need for Capital Discipline, Greater Focus on Operating Cost Control and M&A



Capital discipline: Greater and more explicit awareness of the need for capital discipline is of most pressing importance. In a commoditized market, the only strategic variables in the miners' control are production volume and price, which are inversely related. Expending capital to depress commodity prices is very rarely optimal locally and invariably suboptimal globally.

Focus on operating cost reductions: We would much rather see capital expended to improve productivity at existing operations and effect cost reductions than grow volumes. We believe that geological and technological differentiation is what drives margins. Incumbent players with privileged assets ought to be better able to drive cost reductions than those with more compromised assets. Consequently, the benefits of productivity improvement by the majors should accrue to the shareholders of the incumbents instead of being passed on to consumers through lower prices.

M&A: Keeping the best assets within the portfolios of the incumbent majors is a key barrier to entry for new players. We see significant value (even if it is difficult to quantify) in the optionality inherent in owning the best geology, even if that means expending capital on acquiring assets whose exploitation will be some way in the future. However, the rush to develop assets engendered by too simple an understanding of value creation and the "time value of money" represents a significant flaw in most mining asset valuation methodologies. The big difference between acquisitive and organic growth is that M&A does not add to the supply of raw materials. Therefore, it does not put downward pressure on price. Consequently, to the extent that there is "spare" capital in the miners, in our view, it makes much more sense to buy rather than build.



Grade Is King

Ninety percent of copper prices over the last 35 years can be explained by just three variables: mined grade, global GDP growth and the level of terminal market inventory. In this chapter, we show the mechanism that will continue to drive the grade of mined copper downwards and its price implications. In the previous chapters, we showed that sub-economic resources will increasingly be required to satisfy copper demand going forward, as the financial mechanism behind the accelerated exploitation of existing reserves begins to break down. This chapter analyzes the role that grade plays in copper mining. While copper grade declines are a widely known feature of the mining industry, the true importance of these declines is not. For the first time in the last 35 years, global head grades have moved above reserve grades. Consequently, copper grades are on an inexorable path downwards, and continued exploitation of the ore that stands behind today's copper consumption necessarily degrades the remaining reserve base. While the apparent global copper reserves still stand above 30% of consumption, this has been achieved only through the expedient of slashing of the minimum threshold of copper that is allowed to stand behind reserves (i.e., the cut-off grade). We do not see a way for this situation to be stable in the face of declining copper prices. Again, our analysis leads us to the conclusion that the satisfaction of future demand will require copper price to rise in excess of US$10,000/t.

Three Definitions of Grade It is hard to underestimate the importance of grade in base metal mining. There is an old mining aphorism "grade is king," which captures the importance of grade perfectly. It is the high-grade mines and deposits that continue to make money through the cycle. Consequently, while scale is responsible for the economic viability of a deposit, the grade will determine its cost position and the ultimate value. It is widely known that grades in the copper industry have generally been under pressure. However, the term "grade" covers a multitude of different purposes, and it is important to be clear about what exactly we are referring to when asserting that grades are declining. The implications of deterioration in grades differ markedly depending on what exactly the statement refers to. At the very least, there are three crucial distinctions. Reserve grade is the average grade of the ore that has been identified as

economically viable to extract in the current conditions. So it represents the total available material that the mine will exploit over its operating life.

Head grade is the average grade of ore fed into the processing plant of a mine. For a copper mine, it will either be the mill (for the traditional sulphide route) or the leach pad (for the SxEw route). It refers to the quality of material standing behind today's production and costs.

Cut-off grade is the minimum grade of ore used to establish reserves. It is the threshold grade between the ore that will eventually be fed into the processing plant and the material that will be discarded as waste.

The distinction between the average measures of both head and reserve grades stands in contrast to the limit measure represented by the cut-off grade. This distinction carries significant implications, as we will explain. However, a further explanation of the nature of averages used in measuring grade is necessary. Exhibit 95 shows the differing evolutions of copper grade when a measure is weighted by paid metal (i.e., weighted by the output of the mining process) versus when the same grade is weighted by ore milled (i.e., weighted by the input into the mining process). The difference arises due to a number of very large high-grade mines that

The Operation of a Mine

The Operation of a Mine Depends on the Interplay of the Reserve Grade, the Head Grade and the Cut-Off Grade



account for a disproportionate share of the copper metal output. Were all assets equal in terms of size and grade, the disparity would not have arisen. In the next section, we will be looking at grade weighted by tons of ore milled as it is the best indicator of the material that actually leaves the mine site.

In terms of determining the economics of mine development, the most important measure is the cut-off grade. It will determine the scale of the mining operation, the volume of material that will be exploited and the ultimate life of the mining operation. It also directly influences the quality of ore that is processed on a given day. Consequently, it proves critical in setting the operating cost of a mine. The general rule is that the lower the cut-off grade, the higher the total volume of material available and, consequently, the greater the scale and life of an operation. However, the lower the cut-off grade, the lower the average grade of material and, consequently, the higher the cost of operation. It is the balance between scale and quality that ultimately determines the cost and hence the overall economics of a mine.

Exhibit 95 We Need to Be Careful When Referring to Declines and Understand How Those Declines Are Measured


The Interplay of Grades We begin our analysis by examining how the three types of grades influence the mining process. There will always be some relationship between a deposit's copper grade and some spatial parameters that describe the given ore body (see Exhibit 96). The easiest way to think about this is to assume that the grade declines as one moves down through an ore body. At a certain depth, the copper contained in a volume of rock is simply too low to make the exploitation of that material worthwhile. However, the average grade of the material above this threshold must be considerably higher than the grade that determines the threshold separating ore from waste.

In order to quantify the nature of the relationship between the various mining parameters, it is necessary to specify the grade distribution in some way. To make things easy for ourselves, we have simplified this distribution as an exponentially declining relationship between depth and grade, which we might obtain for a large copper porphyry deposit (see Exhibit 97). Now, an exponential decline is not one

0.60

0.80

1.00

1.20

1.40

1.60

1.80

% C

u

Mined Copper Head Grade

Weighted by Paid Copper Weighted by Ore Milled

Grade Declines as One Moves Down Through an Ore Body, and an Exponentially Declining Relationship Exists Between Depth and Grade for Copper Porphyry Deposits



that could actually be realized in practice. The mineralogical barrier we described in the previous chapters renders a continuous grade distribution impossible. The distribution of ore grades must exhibit a discontinuity for grades below ~0.1% Cu and an exponential function exhibits no such discontinuity. However, the impact of this technical point can be neglected for the time being and the main conclusions drawn from the simplified relationship still stand.

Exhibit 96 The Cut-Off Grade Delineates the Boundary Between the Material That Will Be Mined and Treated and the Waste Material That Will Need to Be Mined to Access the Ore, But Will Be Dumped Subsequently


Direction of Declining Grade

Ore

Waste

Cut‐Off Grade as a Limit

Reserve Grade is the Average of the Economically

Exploitable Ore



Exhibit 97 For Purposes of Illustration, We Assume a Grade Profile That Declines Exponentially With Depth; at Some Depth, Grade Will Be Insufficient to Allow the Material to Be Mined Economically


The lower the cut-off grade is allowed to fall, the more volume of material is opened up for exploitation and, consequently, the larger the mine will be (see Exhibit 98). It is also possible to calculate the relationship between the average grade of all the material to be exploited by looking at the grade of the tons above the cut-off grade. This is measured by the reserve grade of a deposit (see Exhibit 99). The reserves will be extracted over the life of the mine. Exhibit 99 describes a fundamental relationship in how the structure of any mine operates and the nature of the material that stands behind the operation. The important point is the concavity of the curve — as the cut-off grade falls, the reserve grade falls even faster. This is also evident in the ratio between the cut-off grade and the reserve grade (see Exhibit 100). At high reserve grades, the cut-off grade will also be high, and the difference between the marginal ton and the average ton mined is small. However, as the reserve grade falls, the distinction between the marginal ton and the average ton becomes even more pronounced.

0.00%

0.50%

1.00%

1.50%

2.00%

2.50%

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 240 250

Gra

de

of O

re

Depth (meters)

Exponentially Declining Grade With Distance

The Lower the Cut-Off Grade Is Allowed to Fall, the More Volume of Material Is Opened Up for Exploitation, and We Have a Larger Mine; When the Cut-Off Grade Falls, the Reserve Grade Falls Even Faster



Exhibit 98 The Total Tonnage That Will Be Classified as Ore Depends on How Much the Cut-Off Grade Is Allowed to Fall; the Lower the Grade That a Miner Is Prepared to Exploit, the More Tons Are Available


Exhibit 99 The Average Grade of the Tons Mined Will Naturally Be Higher With a Higher Cut-Off Grade, as There Will Be Less Low-Cost Material Present to Dilute the High-Grade Ore


0

20

40

60

80

100

120

140

160

180

200

1.95

%

1.85

%

1.75

%

1.65

%

1.55

%

1.45

%

1.35

%

1.25

%

1.15

%

1.05

%

0.95

%

0.85

%

0.75

%

0.65

%

0.55

%

0.45

%

0.35

%

0.25

%

0.15

%

0.05

%

Ore

To

nn

age

(Arb

itra

ry U

nit

s)

Cut-Off Grade (% Cu in Ore)

Ore Tonnage to Cut-Off Grade Relationship

0.00%

0.50%

1.00%

1.50%

2.00%

2.50%

1.9

5%

1.8

5%

1.7

5%

1.6

5%

1.5

5%

1.4

5%

1.3

5%

1.2

5%

1.1

5%

1.0

5%

0.9

5%

0.8

5%

0.7

5%

0.6

5%

0.5

5%

0.4

5%

0.3

5%

0.2

5%

0.1

5%

0.0

5%

Re

serv

e G

rad

e (A

ver

ag

e G

rad

e o

f Ore

)

Cut-Off Grade

Reserve Grade to Cut-Off Grade Relationship



Exhibit 100 However, the Relationship Between the Average Reserve Grade and the Cut-Off Grade Is Non-Linear; as Reserve Grade Falls, the Miner Is Increasingly "Scraping the Bottom of the Barrel"


The discussion so far serves to illustrate how the structure of a mine is delineated, but not how the mine will actually be developed. The choice of head grade ultimately refers to the "life of mine" plan and how the removal and use of different volumes will be sequenced. This sequencing must, of course, look at more than just grade. It must account for the fact that different grades may have different stripping ratios (requiring one to move more waste to access the ore), and that the geometry of an ore body is seldom simple. In reality, significant computational effort and mine planning software is expended to put together a cost-minimizing (and value maximizing) mine plan. However, the point that we wish to make does not require such complexity. Rather, we wish to explain the implications of having different head and reserve grades. It is possible to develop a mine with no temporal variation in the grade of material exploited. In such case, each year's mining corresponds to taking an identical "slice" out of the ore body. To the extent that this happens, the head grade (the average grade exploited material in a given year) will be identical to the reserve grade, i.e., the material removed in a year and the material left behind will be identical from a grade perspective (see Exhibit 101). However, to the extent that the head grade is higher than the reserve grade, mining leads to a degradation of the ore body. The material that is left behind after a year's production will be of lower quality than the material taken out (see Exhibit 102).

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.9

7%

1.9

2%

1.8

7%

1.8

2%

1.7

7%

1.7

1%

1.6

5%

1.6

0%

1.5

4%

1.4

7%

1.4

1%

1.3

4%

1.2

7%

1.2

0%

1.1

2%

1.0

4%

0.9

5%

0.8

4%

0.7

1%

0.5

3%

Cu

t-O

ff t

o R

ese

rve

Gra

de

Ra

tio

Reserve Grade

Ratio of Cutoff Grade to Reserve Grade With Falling Reserve Grade

The Choice of Head Grade Ultimately Determines the "Life of Mine" Plan and How the Removal and Use of Different Volumes Will Be Sequenced



Exhibit 101 The Mine Plan Determines the Sequencing of Blocks of Material Extracted from the Reserve Base; in This Instance, the Block Taken and the Block Left Behind Are Identical from a Grade Perspective; as a Result, the Mine Plan Sees a Grade Profile That Is Constant Over Time



Ore

Waste


Year 1 Production, a homogeneous slice taken out of the ore body leaving residual ore grade invariant.



Exhibit 102 Under a Different Sequence, the Mine Plan Removes the Highest Grade First, Leaving Lower Grade Material Behind; Over Time, This Means That the Head Grade Must Fall as the Cut-Off Grade Limit Is Approached



Ore

Waste


Year 1 Production, a heterogeneous slice taken out of the ore body leaving residual ore grade lower than

prior to mining.



There are two basic actions that may see an instantaneous head grade higher than the deposit reserve grade. The mine plan may be preferentially extracting high-grade ore zones from the ore

body while leaving lower grade ore in situ. Thus, the head grade of the material taken to the milling plant will be higher than the average grade of the material in the deposit. This allows high grading without increasing the size of the mining fleet.

Low and high-grade ore may be extracted from the mine at the same time, but the low-grade ore will be stockpiled with only the high-grade ore being put through the processing plant. The low-grade ore will be exploited from the stockpile once the mine site itself is exhausted. However, this approach requires an increase in total tons moved, hence an increase in mining fleet.

Irrespective of the mine plan employed, the net result sees head grade above reserve grade in the short run being compensated for by head grade falling below reserve grade (and tending towards the cut-off grade) in the medium to long run. In contrast, if head grades are below reserve grades today, the grades will increase in the future as higher-quality material gets accessed later. Three related questions follow immediately from this analysis: Why should different mine profiles exist, i.e., is there any reason why mine

campaigns preferentially attack low grade first versus high grade first or vice versa?

What are the cost implications behind different grade profiles? Where, in the process of grade development, do we stand today?

Both Geology and Finance Prioritize High-Grade Extraction Maximizing high-grade ore at the start of a mine's life dramatically increases both the value of the mine and the chance of project investment approval. Meanwhile, supergene enrichment provides a geological mechanism that makes high-grade ore available.

A first glance, it might seem odd that there is a systemic reason why high-grade ore rather than low-grade ore should be mined first. After all, it is not as if the nature favors any particular orientation of ore body. Ought one not to be as likely to see low-grade ore close to the surface (and therefore mined first) as high-grade ore?

This is not quite the right picture to have. As is almost always the case in mining, financial and geological reasons exist why high-grade ores are extracted preferentially and low-grade ores subsequently.

To start with financial motivation first, Exhibit 103 shows three different grade profiles for developing exactly the same reserve. The total amount of metal extracted from the mine is identical. Sequencing of extraction is the only variable. Assuming exactly the same capital investment, Exhibit 104 shows the value of each development path (given that the milled tonnage is the same in each scenario, it is fair to assume the same capital costs). High-grading the deposit and extracting cash earlier rather than later has a huge impact on value. In addition to changing the value, it alters the investment decision for building copper projects. High-grading is not just a "sweetener" for investments that would just as well proceed without it. Rather, it is critical for new investment economics.

There Are Two Cases When Head Grade Can Be Higher Than the Reserve Grade: the Mine Plan Is Preferentially Extracting High-Grade Ore Zones First or Low-Grade Ore Is Stockpiled With Only High-Grade Ore Being Processed

From a Financial Perspective, Frontloading Cash Generation Is Preferable Due to the Time Value of Money; from a Geological Perspective, Secondary Enrichment Is Closer to the Surface Than the Low-Grade Ore Zones



While generating the same total lifetime cash, high-grading the ore body delivers cash sooner. Exhibit 105 shows the basic reason for this improvement — the time value of money. Given that the total metal and the total capex are the same under all three scenarios, the value uplift of high-grading is due entirely to it paying back the initial investment sooner.

Nevertheless, this financial incentive is insufficient. After all, if all deposits were perfectly homogenous, high-grading would be impossible. However, again as we discussed in the previous chapters, geology actually conspires to help miners develop projects (see Exhibit 106). Secondary enrichment (whereby weathering on a sulphide outcrop results in the formation of high grade zones where the ore body intersects the historic water table) provides the miners with exactly what they need to maximize the time value of money. However, even without this, the general heterogeneity in a natural ore body will always give clever mine planners the ability to target higher cash flow upfront and leave the painful fallow years for someone else to deal with.

Exhibit 103 Three Different Grade Profiles for the Development of Exactly the Same Underlying Ore Body...


0.6%

0.7%

0.8%

0.9%

1.0%

1.1%

1.2%

1.3%

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

He

ad G

rad

e (%

Cu

)

Years

Three "Life of Mine" Grade Profiles

Scenario 1 - Flat Grade Scenario 2 - Increasing Grade Scenario 3 - Declining Grade



Exhibit 104 ...Lead to Radically Different Value Propositions; Under Realistic Mining Investment Guidelines, Only the Third Scenario of Declining Ore Grade Will Get Approved and Subsequently Developed; the Other Two Mining Solutions Are Unlikely to Attract Capital


Exhibit 105 While Generating the Same Total Lifetime Cash, High-Grading the Ore Body Delivers Cash Sooner


-100

-50

0

50

100

150

200

250

300

350

400

450


NP

V (U

S$m

)

Value Impact of Different Grade Profiles

-2000

-1000

0

1000

2000

3000

4000

5000

6000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

Cu

mu

lati

ve

Pro

ject

FC

F (U

S$m

)

Project Time Line (Years)

Cumulative FCF vs. Mine Plan




Exhibit 106 Geology Conspires With Finance, as Secondary Enrichment Zones Help Create High-Grade Ore Near the Surface, Which Can Be Extracted Early on in the Mine Life


Primary Sulphide Ore

Secondary Enriched Sulphide Ore

Leached or Oxidised Zone

Water Table

Surface

Action of water in dissolving sulphide mineralisation

Supergene Zone

Hypogene Zone



Grade — The Most Important Determinant of Mining Costs Two effects drive the sensitivity of costs to falling grades. First, the same mining and milling expense must cover a smaller output of contained metal in the ore. Second, the efficiency of metal recovery declines as the grade falls. In all of this, "grade is king" and low-cost production is entirely predicated on high grades.

Having understood the structure of grade profiles and the incentive that drives the exploitation of high-grade zones early in the mine plan, we now turn to the impact this has on mine costs. Total cost of a mining operation is driven not by the amount of metal produced, but rather by the amount of ore and waste moved and the volume of ore milled. The ore tonnage scales the total cost base. However, the grade determines the unit costs, as it determines the amount of metal available to bear the total costs of the operation. As a result, the unit costs of a mining operation are inversely proportional to the grade of the material exploited. Four buckets of costs matter to a copper operation. Labor involves paying the people that drive the trucks, run the mills and service

the mine fleet as well as look after the general safety and well-being of the operation.

Diesel is the fuel source for the trucking and mining fleet. Power is used in the crushing of ore in the milling circuits to grind it prior to

froth flotation. Consumables include explosives, tyres, grinding media and reagents used in the

flotation cells. We will be turning our attention to the cost structure of mining more explicitly

in the following chapter, which will look at differential cost escalation and the incentive structure of the industry. However, here we want to examine how various physical factors influence the costs of a mine, even if cost escalation per se (e.g., rising fuel prices) is left out of the equation. Exhibit 107 shows a simplified schematic for the operating cost structure of a copper mine. We deliberately try to mimic the costs of a large Chilean-style copper porphyry mine, exploiting reasonably high-grade ore with an efficient and productive labor force. In Exhibit 108, we show a small-scale, low-efficiency and high-cost mine to illustrate the type of operation that we believe is responsible for the 1Mtpa increase in Chinese copper production over the last decade.

Having constructed this simple model, it is possible to show the sensitivity of operating costs to changes in mining parameters and efficiencies. To pick just one example, the hardness of a rock will determine the power that is required to grind that rock to a suitable size fraction (see Exhibit 114). In general, the sensitivity analysis for a range of parameters is included in Exhibit 109 through Exhibit 115. Exhibit 116 displays the most important factor — different head grades translate to the greatest differences in mining costs.

A second-order effect of falling grade that is often overlooked relates to the separation of valuable copper concentrate from worthless tailing in the flotation cells. Froth flotation is based on the differences in surface chemistry that arise between copper sulphide ores and the gangue of the host rock when treated with certain reagents. However, any process of physical or chemical discrimination requires the differences exploited to be sufficiently wide to induce separation. When the grade of ore falls, the distinction between tailing and concentrate necessarily gets smaller. This, in turn, reduces the ability of flotation cells to discriminate between revenue-generating material and waste product. Consequently, recovery (productivity) falls (see Exhibit 117). Once the cost impact of this second-order effect is included, the role of grade in cost determination becomes even more pronounced (see Exhibit 118). The grade profile of the copper industry is critical for its cost structure and consequently price.

Head Grade Is the Greatest Differentiator in Contemporary Mining Costs; Falling Head Grade Also Has a Second-Order Impact via Reduced Recovery Rates



Exhibit 107 The Basic Structure of the Operating Costs of a Stylized Large Chilean Copper Mine


Simplified Copper Mine Cost Calculation — Low-Cost Mine

Milling Rate ktpd a 100

ROM Tonnage Mtpa b = a * 365 / 1,000 37

Stripping Ratio - c 2.0

Tons Moved Mtpa d =b * (1 + c) 110

Labor Productivity kt/Man-Year e 100

Labor Required Men f = d / e * 1,000 1,095

Unit Labor Cost US$pa g 100,000

Total Labor Cost US$m h = f * g / 1,000,000 110

Diesel Consumed Liters/t Moved i 0.75

Diesel Price US$/Liter k 1.00

Unit Diesel Cost US$/t Moved l = i * k 0.75

Diesel Cost US$m m = l * d 82

Unit Mining Consumables Cost US$/t Moved n 1.00

Mining Consumables Cost US$m o = n * d 110

Total Mining Cost US$m p = o + m + h 301

Milling Power Intensity kwh/t Milled q 25.0

Unit Power Cost USc/kwh r 7.5

Milling Power Cost US$m s = b * q * r / 1,000 68

Milling Consumables Cost US$/t Milled t 2.0

Milling Consumables Cost US$m u = t * b 73

Milling Labor Productivity kt/Man-Year v 55

Milling Labor Required Men w = b / v * 1000 664

Milling Labor Cost US$m x = w * g / 1,000,000 66

Total Milling Cost US$m y = s + u + x 208

Milling & Mining Cost US$m z = y + p 509

G&A as % of Total Costs % aa 20%

Other G&A US$m ab = aa / (1-aa) * z 127

Total Minesite Cost US$m ac = ab + z 636

Head Grade % ad 1.00%

Milling Recovery % ae 85%

Concentrate Grade % af 35%

Concentrate Produced ktpa ag = b * ad * ae / af *1000 886

Recovered Metal ktpa ah = ag * af 310

Metal Payability % ai 97.5%

Payable Metal ktpa aj = ah * ai 302

Land Freight US$/t Concentrate ak 20

Ocean Freight US$/t Concentrate al 50

Freight Cost US$m am = ag * (ak + al) / 1,000 62

TC US$/t Concentrate an 80

RC Usc/lb Copper ao 8

Treatment Charge US$m ap = an * ag / 1,000 71

Refining Charge US$m aq = (2,204 * ao * aj) / 10,0000 53

Realization Cost US$m ar = ap + aq 124

Total Cost US$m as = ac + ar + am 822

Cost per Ton US$/t at = as / aj * 1000 2,719



Exhibit 108 The Picture for a Small Chinese Type Operation (or at Least What We Believe They Look Like!) Is a Little Different


Simplified Copper Mine Cost Calculation — High-Cost Mine

Milling Rate ktpd a 1

ROM Tonnage Mtpa b = a * 365 / 1,000 0.37

Stripping Ratio - c 3.0

Tons Moved Mtpa d =b * (1 + c) 1.46

Labor Productivity kt/Man-Year e 10

Labor Required Men f = d / e * 1,000 146

Unit Labor Cost US$pa g 8,000

Total Labor Cost US$m h = f * g / 1,000,000 1.17

Diesel Consumed Liters/t Moved i 0.75

Diesel Price US$/Liter k 1.00

Unit Diesel Cost US$/t Moved l = i * k 0.75

Diesel Cost US$m m = l * d 1.10

Unit Mining Consumables Cost US$/t Moved n 1.00

Mining Consumables Cost US$m o = n * d 1.46

Total Mining Cost US$m p = o + m + h 3.72

Milling Power Intensity kwh/t Milled q 25.0

Unit Power Cost USc/kwh r 7.5

Milling Power Cost US$m s = b * q * r / 1,000 0.68

Milling Consumables Cost US$/t Milled t 2.0

Milling Consumables Cost US$m u = t * b 0.73

Milling Labor Productivity kt/Man-Year v 10

Milling Labor Required Men w = b / v * 1000 37

Milling Labor Cost US$m x = w * g / 1,000,000 0.29

Total Milling Cost US$m y = s + u + x 1.71

Milling & Mining Cost US$m z = y + p 5.43

G&A as % of Total Costs % aa 20%

Other G&A US$m ab = aa / (1-aa) * z 1.36

Total Minesite Cost US$m ac = ab + z 6.79

Head Grade % ad 0.30%

Milling Recovery % ae 85%

Concentrate Grade % af 35%

Concentrate Produced ktpa ag = b * ad * ae / af *1000 2.66

Recovered Metal ktpa ah = ag * af 0.93

Metal Payability % ai 97.5%

Payable Metal ktpa aj = ah * ai 0.91

Land Freight US$/t Concentrate ak 20

Ocean Freight US$/t Concentrate al 0

Freight Cost US$m am = ag * (ak + al) / 1,000 0.05

TC US$/t Concentrate an 80

RC Usc/lb Copper ao 8

Treatment Charge US$m ap = an * ag / 1,000 0.21

Refining Charge US$m aq = (2,204 * ao * aj) / 10,0000 0.16

Realization Cost US$m ar = ap + aq 0.37

Total Cost US$m as = ac + ar + am 7.21

Cost per Ton US$/t at = as / aj * 1000 7,948



Exhibit 109 Stripping Ratio: The Deeper You Go the Higher the Cost

Exhibit 110 Labor Productivity: Efficient and Well-Capitalized Mines Are Far Lower Cost Than Those Relying on Human Labor

Source: Bernstein estimates and analysis. Source: Bernstein estimates and analysis.

Exhibit 111 The Price of Labor Matters... Exhibit 112 ...While the Impact of Diesel Is Surprisingly Low


1,000

1,500

2,000

2,500

3,000

3,500

US

$/t

Stripping Ratio

Cost Sensitivity to Stripping

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

5,000

5,500

US

$/t

kt/man-year

Cost Sensitivity to Labor Productivity

1,000

1,500

2,000

2,500

3,000

3,500

US

$/t

US$/man-year

Cost Sensitivity to Labor Cost

1,000

1,200

1,400

1,600

1,800

2,000

2,200

2,400

2,600

2,800

3,000

US

$/t

US$/liter

Cost Sensitivity to Diesel Price



Exhibit 113 Power Has Less Impact Than One Might Expect

Exhibit 114 Rock Hardness Affects the Milling Power Rate and Drives Cost in High Power-Cost Regions


Exhibit 115 Higher Mill Recovery (Finer Grind) Implies Lower Cost

Exhibit 116 Nevertheless, in All of This "Grade Is King"


1,000

1,500

2,000

2,500

3,000

3,500

US

$/t

USc/kwh

Cost Sensitivity to Power Price

1,000

1,200

1,400

1,600

1,800

2,000

2,200

2,400

2,600

2,800

3,000

US

$/t

kwh/t milled

Cost Sensitivity to Milling Power Rate

1,000

1,200

1,400

1,600

1,800

2,000

2,200

2,400

2,600

2,800

3,000

US

$/t

Mill Recovery

Cost Sensitivity to Mill Recovery

1,000

3,000

5,000

7,000

9,000

11,000

13,000

15,000

17,000

US

$/t

Head Grade

Cost Sensitivity to Head Grade



Exhibit 117 Compounding the Critical Role That Grade Plays Is the Fact That the Productivity of Milling Circuits Varies With It; a "Constant Tail" of Copper Is Always Lost, Given That Below a Certain Copper Concentration, Froth Flotation Cannot Discriminate Between Ore and Tailings


Exhibit 118 This Radically Increases the Cost of Low-Grade Volume Exploitation


0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0.00% 0.10% 0.20% 0.30% 0.40% 0.50% 0.60% 0.70% 0.80% 0.90% 1.00% 1.10% 1.20% 1.30% 1.40%

Mil

l Re

cov

ery

Head Grade (% Cu)

Mill Recovery vs. Head Grade

2,000

4,000

6,000

8,000

10,000

12,000

14,000

0.20% 0.30% 0.40% 0.50% 0.60% 0.70% 0.80% 0.90% 1.00% 1.10% 1.20%

Co

pp

er

Co

sts

(US

$/t

)

Head Grade (% Cu)

Impact of Mill Recovery on Cost Sensitvity of Grade Decline

Including Recovery Impact Excluding Recovery Impact



Looking Beyond the Superficial Today, global copper reserves have a life of roughly 30 years, and this is relatively unchanged from 35 years ago. However, the persistently steady reserves have only been made possible through a dramatic decline in cut-off grades. Apparent reserve security belies the radical change implied in moving from a situation of deposit head grades being below reserve grades to one where the reverse is true. As a result, the security of 30 years worth of reserves would vanish in the face of a dramatic price decline.

So far, we have discussed three issues: Established the nature of the grade relationships in copper mining Explained why the grade profiles drive mine life plans Demonstrated the importance of grade to the global cost structure

We now turn to the final piece of the puzzle and look at the current state of the global copper industry, in order to determine the implications of all these considerations on the copper price. At first glance, the overall picture for the security of copper supply considered from a global perspective does not look very different now compared to 35 years ago. At least on a very superficial level, one would look at Exhibit 119 and conclude that despite the fact that mined copper production has grown by nearly 10Mtpa, the world has been able to meet this increase, and that any Malthusian concerns about imminent supply shortage are overdone (see Exhibit 120).

The driving force behind this is the fact that copper reserve base has grown more or less in proportion to production, leading to only modest declines in global mine life (see Exhibit 121). However, this is not the complete story. In fact, the truth is far more complicated. The superficial statements that there is plenty of copper left in the ground and our ability to replace what we consume means that production is secure would be true only in so far as price stays high and trends higher with time. In fact, reserves have only grown together with increasing consumption because the global reserve grade has collapsed (see Exhibit 122). Reserves have held up only because resources have been converted into reserves by dramatically lowering the cut-off grade (see Exhibit 123).

Exhibit 119 The Life of Available Copper Deposits Has Held Up Very Well, Despite the Massive Increases in Mined Production

Exhibit 120 So There Is No Need for Any Malthusian Concerns; We Have as Much Copper "on Hand" Now as 35 Years Ago


0

5

10

15

20

25

30

35

40

45

1980 1985 1990 1995 2000 2005 2010

Ye

ars

Global Copper Reserve Life

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

1980 1985 1990 1995 2000 2005 2010

Co

pp

er (k

t)

Global Mined Copper Production

The Reason Why Reserves Have Held Up Is Because Resources Have Been Converted Into Reserves by Dramatically Lowering the Cut-Off Grade; Hence, 30 Years Worth of Reserves Would Vanish in the Face of a Dramatic Price Decline



Exhibit 121 This Security Is Engendered by the Fact That Reserves Seem to Have Held Up Well

Exhibit 122 However, the Truth Is Somewhat More Complicated; Reserves Have Held Up Well Only Because of a Dramatic Reduction in the Reserve Grade...


Exhibit 123 ...Driven by Dropping the Cut-Off Grade to below 0.3% Cu


The inadequacy of reserves under current economic conditions necessitates changing the conditions that are deemed necessary. The geology is invariant, the economics must change. However, although the declines in global head grade (i.e., the material that stands behind today's cost structure and copper price) may be well known, the implications of this for future copper prices are not (see Exhibit 124). It is the relative decline in the prevailing head grade, the reserve grade and the cut-off grade that matters for price. Critically, we see that 10 or so years ago, global head grades moved from being below global reserve grades to being significantly above them (see Exhibit 125 and Exhibit 126). This move reflects a change in global mining practice. Twenty years ago, the extraction of copper in a given year did not degrade the quality of the ore body from which the copper was extracted. In fact,

0

100

200

300

400

500

600

1980 1985 1990 1995 2000 2005 2010

Co

nta

ined

Cu

(Mt)

Global Copper Reserves

0.40%

0.50%

0.60%

0.70%

0.80%

0.90%

1.00%

1.10%

1980 1985 1990 1995 2000 2005 2010

Av

era

ge

Res

erv

e G

rad

e (%

)

Global Copper Reserve Grade

0.20%

0.25%

0.30%

0.35%

0.40%

0.45%

0.50%

0.55%

1980 1985 1990 1995 2000 2005 2010

Av

era

ge

Res

erv

e G

rad

e (%

)

Global Copper Cutoff Grade

Extraction of Today's Ore Has a Deleterious Impact on the Quality of Ore That Will Be Left to Satisfy Demand Tomorrow



the average quality of the ore left behind was slightly higher than the ore extracted, as below-average grade material was taken out first. The reverse is true now: the highest-quality fractions of the global reserve base are required to satisfy demand, and the extraction of today's ore has a deleterious impact on the ore that will be left for tomorrow (see Exhibit 127 and Exhibit 128). The transition in the copper industry over the last 10-15 years has been from self-perpetuating to one that is increasingly unstable in the face of price shocks or other disruptions.

Exhibit 124 While the Decline in Head Grade Is Well Known...

Exhibit 125 ...Its Implications Are Not, Given the Nature of the Copper Forward Curve and Consensus Price Expectations


Exhibit 126 Production Over the Last Decade Has Been Sustained Only Through Moving Head Grades Above Reserve Grade and Global High-Grading


0.60%

0.65%

0.70%

0.75%

0.80%

0.85%

0.90%

0.95%

1.00%

1980 1985 1990 1995 2000 2005 2010

Hea

d G

rad

e (%

)

Global Copper Head Grade

0.0%

0.2%

0.4%

0.6%

0.8%

1.0%

1.2%

1980 1985 1990 1995 2000 2005 2010

% C

u

Reserve to Head Grade

Reserve Grade Head Grade

0.8

0.9

1.0

1.1

1.2

1.3

1.4

1980 1985 1990 1995 2000 2005 2010

Head Grade to Reserve Grade Ratio



Exhibit 127 This Means That Grades Must Decline Sharply in the Future; in Fact, Grades Must Decline Towards the Much Lower Level Given by the Cut-Off Grade


Exhibit 128 2003 (or Thereabouts) Saw a Structural Break in the Industry, Which Moved from a Period Where Mining Did Not Imply Geological Degradation to One Where It Most Certainly Does


0.3%

0.4%

0.5%

0.6%

0.7%

0.8%

0.9%

1.0%

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

Hea

d G

rad

e F

ore

cast

(%C

u)

Years Out from Forecast Date

Implied Head Grade With Exponential Decline

1980 1985 1990 1995 2000 2005 2010

30

40

50

60

70

80

90

100

110

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30Ind

exe

d H

ead

Gra

de

Rel

ativ

e to

In

itia

l Le

vels

Years Out from Forecast Date

Indexed Implied Head Grade With Exponential Decline

1980 1985 1990 1995 2000 2005 2010



In Exhibit 129, we show the long-term history of Australian copper grades. It is remarkable that little more than a century and a half of industrial copper consumption has seen the grade fall from well in excess of 20% to below 1%. A belief that this trajectory is commensurate with ever declining real copper prices seems, at least to us, absurd. However that may be, the more serious point is illustrated in just the last century of Australia's development (see Exhibit 130). Australian copper production has stagnated — it has hit the threshold we described in the previous chapters, wherein the financial barrier to easy expansion caps the maximum production that a given resource base can support. However, production has not declined. Instead, grade has declined by over 62%, as the material that was previously uneconomic to mine has been dragged in to support production. There is no way that this would have happened, had copper price increases not allowed it. Once again, the economics of copper mining had to change to enable the "reserve to resource" conversion, once simple mine expansion was no longer viable.

Exhibit 129 Just for Fun, We Provide a Long-Term History of Copper Grades in Australia

Source: Monash University and Bernstein estimates and analysis.

0

5

10

15

20

25

30

Co

pp

er H

ead

Gra

de

-% C

u

Australian Copper Grade Over Time

A Century and a Half of Industrial Copper Consumption Has Seen Grade Fall from >20% to <1%, Which in Itself Makes It Obvious That the Prospect of Ever Declining Real Copper Prices Is Absurd



Exhibit 130 But on a Serious Point: Post-2000, We See Stagnation in the Copper Industry, Which Saw Mine Output Holding Up Only Through a Radical Grade Decline; What Happened in Australia Was Repeated Globally

Source: Monash University and Bernstein estimates and analysis.

Three Things You Need to Know About Copper The relationship between head grade and cut-off grade enables one to calculate the trajectory of future head grades (see Exhibit 131). Because head grades are very high today, they must fall tomorrow. Our analysis suggests that the risk of grades falling faster than many anticipate is high. However, any grade decline (let alone a more rapid one than may be embedded in consensus price expectations) would result in a radical change in the cost structure of the copper mining industry. We cannot see how this change in cost structure can be balanced with an increased demand environment and a falling or flat price dynamic.

Fortunately, one can do more than just qualitatively state the direction of the impact. In Exhibit 132 and Exhibit 133, we show how prices and grades track each other. This immediately suggests the structural form for a multivariate regression that explains (without the introduction of specious exogenous regime shifts and Heaviside functions) the copper price evolution over the last 35 years.

The following regression explains nearly 90% of the variation in copper price over the last 35 years.

Price = + 1*Grade +2*GDP +3*Inventory The elements in this are not a result of data mining with an aim to find the

highest R-squared. Rather, they are included for their economic significance. All three variables serve as a proxy for an economic determinant of price. In particular, they proxy the cost structure of the industry (grade), global demand (GDP) and global supply (inventory).

By far, the most important driver of the three is grade. This relationship gives us the effective real-world cost increase implied in grade decline. The sensitivity analysis in the previous section illustrated the theoretical impact of grade on cost and, ultimately, price. However, it was necessarily a simplification of the multitude of factors at play in the real world. The regression analysis enables one to see the actual impact of falling grades (see Exhibit 134 and Exhibit 135).

-

100

200

300

400

500

600

700

800

900

1,000

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

7.0%

8.0%

1900

1905

1910

1915

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

Au

stra

lian

Co

pp

er P

rod

uct

ion

-kt

Au

stra

lian

Co

pp

er G

rad

e -%

Grade and Production Relationship in Australia

Cu Grade Cu Production

A Multivariable Regression With Three Variables (Grade, Global GDP Growth and LME Inventory) Explains 90% of the Variation in Copper Price Over the Last 35 Years; It Is Only a Matter of When, Not If, Copper Prices Will Exceed US$10,000/t



We also show the distribution of the errors between the fitted copper price and the actual copper price, in order to illustrate that the errors are normally distributed, so there is no evidence of bias that an otherwise overlooked explanatory variable would need to correct for. In our view, the error between the actual and predicted copper price is just "noise" rather than a reflection of any structural oversight (see Exhibit 136). We also include the statistical summary of the relationship in Exhibit 137. This is important because it highlights the p-values of the fit, which represent a measure of chance that the regression relationship was due to a coincidence. All of the regression parameters are significant at well below the 1% level. Collectively, the regression has a p-value of 1.3E-16. In other words, the regression has a 1 in 10 quadrillion (thousand trillion) chance of being the result of luck alone.

Consensus copper price forecast is essentially flat with a slight longer-term backwardation. The strength of the regression poses a very strong challenge to the consensus view. Our disbelief in the consensus forecast comes from the inescapable reality of grade declines driven by the current mining practices in the copper industry (see Exhibit 138). We see no way to escape this phenomenon in the medium term. Consequently, any explanation of a copper price that differs from our price target of US$10,000/t must look at something other than grade.

The two most likely variables other than cost are supply and demand (or in terms of our regression analysis — their proxies of global GDP growth and terminal market inventories). These are the only places wherein the bear thesis for this metal can hide. We calculate the sensitivity of copper price to changes in each of these variables (see Exhibit 139 through Exhibit 141).

As expected, grade is by far the most important variable. Differences in supply and demand determine only where on that cost structure price must fall to ensure equilibrium. As we have pointed out before, the price of commodities and the "super-cycle" is primarily a supply rather than a demand-side phenomenon. The sensitivity analysis enables us to reconstruct the implicit (and assuredly not explicit) assumptions discounted by the current copper consensus price.

The most implausible explanation for flat or declining real copper prices is based on demand-side fears. While the stocking and destocking cycle for any metal is a cause of huge volatility, it is simply insufficient to derail copper price appreciation in the face of declining grades. Exhibit 142 shows the global growth trajectory implicit in consensus price expectations. It requires an outcome that would dwarf the global financial crisis of 2008-09 in terms of severity. Clearly, the data points that are described in the exhibit lie outside the range of the regression. This only highlights the unlikely nature of the required demand-side failure. Simply put, we have never seen the kind of growth required to support the bear thesis for the copper price.

Low copper prices could also be explained by oversupply from the miners. Indeed, we are currently at a subdued price for copper as a consequence of such a situation. Metal that does not have a natural industrial demand ends up in terminal market warehouses rather than being embedded into the infrastructure of the global economy in the form of wiring and pipes. But how bad would the oversupply have to be to justify the consensus price path (see Exhibit 143)? The miners would have to continue producing metal that was palpably not required by any consumer on an indefinite basis, leading to the accumulation of unwanted metal in warehouses some 116% higher than the previous historical high of 2002 (the very bottom of the copper cycle).

Even if the level of global growth does not improve and even if metal inventories stay at their current elevated levels (despite the fact that the last few months have seen falling inventories), the copper price would hit US$10,000/t by 2017. In fact, under a range of more plausible growth and grade scenarios US$10,000/t is unavoidable on an even shorter time frame (see Exhibit 144). We strongly believe that the question is when, not if. In order for the consensus price forecast to hold true (at the same time respecting the strength of the explanation of the simple regression presented here), a highly implausible set of supply and



demand circumstances must persist almost without break and without response from either the miners or global policymakers. Hence, the question becomes: "Which is more unlikely, that the regression explaining the entirety of copper price history over the last 35 years stopped working or that the implicit assumptions standing behind consensus copper price expectations are untenable?" Surely, an unbiased interpretation of probability and risk would conclude that it is copper prices that will rise rather than the assumptions discounted by consensus come to pass.

It should also be noted that this analysis excludes the impact of any other real-term cost escalation in the drivers of copper prices (such as rising labor costs in China or increasingly expensive diesel and power). The covariance of the variables with the copper price itself renders them unsuitable for inclusion in the regression analysis discussed earlier.

Exhibit 131 The Relationship Between Cut-Off Grade, Head Grade and Reserve Grade Over the Life of the Reserve Base Enables One to Calculate How Head Grades Will Need to Evolve Going Forward


0.50%

0.55%

0.60%

0.65%

0.70%

0.75%

0.80%

0.85%

0.90%

0.95%

1.00%

1980 1985 1990 1995 2000 2005 2010 2015 2020 2025

Hea

d G

rad

e

Impact of Analysis vs. Consensus Grade Expectations

Wood Mac Forecast Implied Convergence to Cutoff Grade Trend 2003-2013



Exhibit 132 How Is All This Related to Price? Well, in the First Instance, We Can Observe That the Step-Up in Prices Occurred When Global Head Grades Started to Fall


Exhibit 133 Indeed, There Is a Very Strong Relationship Between These Two Factors


0.50%

0.55%

0.60%

0.65%

0.70%

0.75%

0.80%

0.85%

0.90%

0.95%

1.00%

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

10,000

He

ad G

rad

e (%

Cu

)

Re

al C

op

pe

r P

ric

e (U

S$

/t)

History of Copper Grade and Copper Price

Real Copper Price Global Copper Head Grade

R² = 0.8173

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

10,000

0.65% 0.70% 0.75% 0.80% 0.85% 0.90% 0.95% 1.00%

Rea

l Co

pp

er

Pri

ce (U

S$/

t)

Copper Head Grade

Copper Grade to Price



Exhibit 134 We Use This as the Basis of a Simple Multivariate Regression, Which Almost Perfectly Describes the Copper Price


Exhibit 135 Nearly 90% of the Movement in Copper Is Given by Just Three Factors: Grade, Global GDP Growth and LME Inventory


0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

10,000

Rea

l C

u P

rice

(US

$/t

)Fitted vs. Actual Copper Price

Actual Price Fitted Price

R² = 0.8864

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

10,000

0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 10,000

Fit

ted

Co

pp

er

Pri

ce

(US

$/t)

Actual Copper Price (US$/t)

Copper Price Regression



Exhibit 136 The Normalcy of the Errors Between Regressed and Actual Copper Prices Implies No Systemic Bias or Neglected Explanatory Variable…


Exhibit 137 …Leading to a Regression Analysis Whose Component Parts Are All Highly Significant


0

1

2

3

4

5

6

7

8

-1250 -1000 -750 -500 -250 0 250 500 750 1000 1250

Co

un

t of

Err

ors

Error — Actual vs. Regression (US$/t)

Analysis of Errors of Regression

Regression Statistics

Multiple R 94.5%

R-Square 89.3%

Adjusted R-Square 88.2%

Standard Error 699

Observations 33

Coefficients t Stat P-value

Intercept 26,751 17.0 1.2E-16

Head Grade -2,551,983 -14.9 3.8E-15

Global GDP 26,680 2.9 7.5E-03

Termonal Inventory -1.19 -3.0 5.6E-03



Exhibit 138 Grade Declines Are An Irreversible Feature of the Industry, Driven by the Structure of Current Mine Plans That Stand Behind Current Copper Consumption


Exhibit 139 Grade Is by Far the Most Important to Price Sensitivity…


0.00%

0.10%

0.20%

0.30%

0.40%

0.50%

0.60%

0.70%

0.80%

Cutoff Grade Reserve Grade Head Grade

Co

pp

er G

rad

e

Global Average Grade Profile

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

Co

pp

er P

rice

(U

S$/

t)

Head Grade

Price Sensitivty of Regression Analysis to Grade



Exhibit 140 …But Economic Growth Prospects Can Swing the Price by ~US$1,500/t…


Exhibit 141 …With a Similar Order of Magnitude Contribution from Terminal Market Inventory


4,000

4,200

4,400

4,600

4,800

5,000

5,200

5,400

5,600

5,800

4.5

0%

4.2

5%

4.0

0%

3.7

5%

3.5

0%

3.2

5%

3.0

0%

2.7

5%

2.5

0%

2.2

5%

2.0

0%

1.7

5%

1.5

0%

1.2

5%

1.0

0%

0.7

5%

0.5

0%

0.2

5%

Co

pp

er

Pri

ce (U

S$

/t)

Global Economic Growth

Price Sensitivity of Regression Analysis to Global GDP

4,200

4,400

4,600

4,800

5,000

5,200

5,400

5,600

5,800

6,000

Co

pp

er P

rice

(U

S$/

t)

LME, SHFE, COMEX Copper Inventory (kt)

Price Sensitivity of Regression Analysis to Copper Stocks



Exhibit 142 This Implies That GDP Has to "Work" Incredibly Hard to Overcome the Inevitable Grade Decline and Return Consensus Price Expectations; to Put It Mildly, the Implicit Assumption of Consensus Is Implausible…


Exhibit 143 …Likewise for Terminal Market Inventory; Surely, the Most Likely Read Through Is Not That the Miners Forget How to Run Their Businesses, But That Prices Will Trend Higher Than Consensus Currently Anticipates


-8.0%

-6.0%

-4.0%

-2.0%

0.0%

2.0%

4.0%

6.0%

1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015E 2017E

Glo

bal

GD

P G

row

th

Global Growth Required for Consensus Copper Price

GDP Actual GDP Required for Consensus Cu Price Forecast

0

500

1,000

1,500

2,000

2,500

3,000

1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015E 2017E

Ter

min

al In

ven

tory

(kt

Cu

)

Terminal Inventory Required for Consensus Copper Price

Inventory Actual Inventory Required for Consensus Cu Price Forecast



Exhibit 144 Grade Is by Far the Strongest Driver of Price; Under Any Scenario, Dropping the Cut-Off Grade to the Point Where the Head Grade Would Be Below the Reserve Grade Implies Prices Well in Excess of US$10,000/t


0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

198

1

198

3

198

5

198

7

198

9

199

1

199

3

199

5

199

7

199

9

200

1

200

3

200

5

200

7

200

9

201

1

201

3

201

5

201

7

201

9

202

1

202

3

202

5

Rea

l Co

pp

er

Pri

ce (U

S$/

t)

Copper Price Forecasts Under Three Grade Scenarios

WM Head Grade SCB Head Grade Historical Head Grade Trend

Actual Price Trend Price Line



From Grade to Grunt and the Real Impact of Wage Inflation

Margin in mining is generated through geological and technological differentiation. In other words, it depends on both the quality of geology and the methods employed to exploit that geology economically. Having discussed the impact of that declining geological quality of new copper deposits is expected to have on the copper price going forward, we now turn to the second mechanism behind margin generation — the productivity and cost required to develop a particular geological formation. We show that the implicit assumptions behind the current consensus copper price line are, at best, highly unlikely to come true. Specifically, for consensus to prove correct, we would have to see nominal wage declines compensate for the cost effects of geological deterioration. Instead, our analysis suggests that the greatest real-term cost increases will occur in what is already the highest-cost supply location (China). This will inexorably lead to a steepening rather than a flattening of the global copper cost curve. In addition, a close look at the incentive price structure of the copper industry explains why higher prices are unlikely to induce a wave of new supply that could otherwise undermine this structural dynamic.

Mining Margin Generation Inflection in the cost curve of a commodity industry generates the margin for incumbent players. A steep cost curve implies a price that is high relative to the average industry costs, and hence a high margin for the low-cost producers. A flat cost curve implies a price that is low relative to the average costs and, consequently, a low margin for the industry. However, the cost curve structure of an industry is not fixed and varies over time. First, it changes as old mines become ever higher cost and need to access marginal tons at an ever greater depth. Second, it varies as new lower-cost mines are brought on line to displace the higher-cost sources of supply. The resilience of the cost curve inflection to supply side disruptions generates a sustainable margin and creates a strategically attractive industry.

An attractive industry is one where the barriers to entry ensure that higher commodity prices will not automatically translate into unwanted new capacity. In contrast, a strategically unattractive industry has a natural tendency towards cost curve flattening due to an inherent ubiquity in geology and mining method. Strategy in mining is all about one thing — the dynamics of commodity cost curves and the miners' capital allocation choices. These two factors determine the dynamics and simultaneously seek to exploit them. At the simplest level, cost curve inflection depends on two factors. Geological differentiation: A high degree of geological differentiation speaks to

the degree of scarcity behind commodity production (after all, if the commodity was not scarce everyone would be able to find high-grade deposits!). A radical difference in cost is determined by the deposit's grade. Other geological factors such as stripping ratio and rock hardness add further complexity to the issue. Clearly, the best position for a miner is to be a sole possessor of the only high-grade deposit of a mineral in a world dominated by low-grade production.

Technological differentiation: There are several ways in which an ore body can be exploited. Deciding which mine plan is the most value accretive requires skill in project evaluation. The trade-off between capital and operating cost is a

Expected Chinese Mining Wage Inflation and a Lack of Incentive to Sink in New Capex in Copper Projects Today Stand Behind Our Bullish Long-Term Outlook for Copper

Margin Is Generated Through Inflection in the Cash Cost Curve; in Mining, This Requires Either (or Preferably Both) Geological Differentiation or Differences in Exploitation Productivity



fundamental choice in a mine plan. It, in turn, reflects a more basic interplay between the productivity of labor and capital.

These two factors together determine the attractiveness of a particular commodity over the long term. In a world of perfect geological and technological homogeneity, the cost structure of every agent involved in production would be identical, and the cost curve would be perfectly flat. In terms of value creation, a world dominated by identically efficient mines is just as bad as a world dominated by identically inefficient mines. It is the relative difference between the good and the bad mines (rather than the absolute level per se) that generates margin. After all, no one values the commonplace.

Deconstructing Copper Mining and Milling Costs The subsequent analysis exposes some of the unreasonable assumptions implicit in the current consensus expectation of flat or declining commodity prices. We decompose copper mining costs on a region-by-region basis and show how differences in labor costs will likely lead to a steepening in the copper cost curve. We expect this effect to deliver copper prices of US$10,000/t over the medium term. Furthermore, we derive the fully loaded (i.e., capturing the full impact of political and country risk for each mining jurisdiction) incentive price curve for the copper industry. Again, we show why US$10,000/t will be required to close any future supply/demand imbalances.

Previously, we have focused on a generalized cost analysis for copper mines in order to test the copper mining cost sensitivity to various scenarios. In Exhibit 145 through Exhibit 148, we present an overall breakdown of costs for a copper mine by stage of the mining process and by cost category. One should bear in mind that each mine is unique and, consequently, there are very significant variations from one mine site to another. Furthermore, each country will have a unique cost structure that will reflect the local economic circumstances. We have broken down the mining process into three stages: Mining involves the movement of ore and waste material with an aim to deliver

valuable ore to the milling circuits for treatment. Its cost is fundamentally tied to the efficient movement of material in bulk.

Milling is the process of ore crushing and grinding, followed by the separation of valuable copper containing minerals from waste tailings.

G&A relates to the overhead costs of the mining operation. These then yield five cost categories:

Labor is the cost of employing the people working directly at the mine. Services comprise the secondary activities that, while not directly connected to

the mining process, are essential for its ongoing activity (e.g., site security, laboratory work or the canteen). This represents just another form of labor cost.

Diesel is the fuel source behind the mining fleet of trucks and shovels. In addition, it is also required for crushing and grinding. For some operations electrification of the mining fleet is an option. In general, diesel costs will be a very significant part of the mining operation costs, but relatively insignificant for milling.

Power is the electricity used in crushing, grinding and separation of ore. Consequently, it is an important, but a relatively small, part of the milling process.

Consumables cover explosives, tires and lubricants used on the mine site. For the milling step, it will cover grinding media used in the mills as well as chemical reagents used during froth floatation where copper concentrate is separated from gangue tailings.

By Decomposing Mining and Milling Costs We Explore Cost Sensitivity of Copper Mining to Different Parameters and Determine the Differences That Will Translate Into Margin



Exhibit 147 Overhead Costs of a Mine Site Mainly Comprise Labor

Exhibit 148 The Split Between Mining and Milling Is Roughly Even With G&A Being a Relatively Small Part of the Overall Costs

Source: Wood Mackenzie and Bernstein analysis. Source: Wood Mackenzie and Bernstein analysis.

In Exhibit 149 through Exhibit 151, we show the current status of the three main cost categories: diesel, power and labor. We break down our analysis by copper-producing region and highlight the cost positions of top 10 supply locations. Chile is no longer a low-cost mining jurisdiction. In addition, the developed world is a high-cost mining location. Latin America is already a medium- to high-cost producer, and the only low-cost producing regions are Africa and China.

However, of the three main cost drivers, two (diesel and power) show very little geographical variation (see Exhibit 152). This is as expected — oil, coal and gas trade in large measure on global markets with internationally determined prices. On the other hand, there is very significant international variation in the price of labor. Consequently, labor cost evolution has the greatest ability to drive

Labor, 76%

Services,10%

Diesel, 1%

Power, 3%

Consumables, 9%

G&A Cost Composition

Mine, 46%

Mill, 41%

G&A, 14%

Copper Cost Breakdown by Process

Exhibit 145 Labor (Including Services) and Consumables Represent the Largest Cost Elements for the Mine Site

Exhibit 146 Consumables and Power Form the Largest Cost Element for Milling


The Three Main Cost Drivers Are Diesel, Power and Labor, With the Former Two Showing Very Little Geographical Variation Due to the Global Nature of Oil, Coal and Gas Markets

Labor26%

Services21%

Diesel20%

Power5%

Consumables28%

Mine Cost Composition

Labor12%

Services16%

Diesel2%

Power32%

Consumables38%

Mill Cost Composition



differential mining cost escalation, therefore a margin-generating rotation in the cost curve.

Exhibit 150 Chile Stands Out as a Very High Power Cost Region; in Africa, It Is Not the Cost But Rather the Reliability of Power Supply That Is the Issue


0

2

4

6

8

10

12

14

16

18

20

Ma

urita

nia

Chi

le

Arm

enia

Pol

and

Mor

occ

o

Phi

lippi

nes

Por

tug

al

Ind

ia

Lao

s

Mo

ngol

ia

Arg

entin

a

Bra

zil

Me

xico

Pa

pua

Ne

w G

uin

ea

Sp

ain

Aus

tral

ia

Bul

garia

Vie

tnam

Tur

key

Nam

ibia

Ind

one

sia

Chi

na

Per

u

Can

ada

Bot

swa

na

Fin

lan

d

Rus

sia

n F

eder

atio

n

US

A

So

uth

Afr

ica

Bol

ivia

Za

mbi

a

Ira

n

Con

go D

.R.

Sau

di A

rabi

a

Sw

ede

n

Uzb

ek

ista

n

Ka

zakh

stan

Po

wer

Pri

ce (U

Sc/

kwh

)

Power Price by Copper-Producing Country

Exhibit 149 The Price of Diesel Shows Relatively Small Variation from One Mining Jurisdiction to the Next


0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

Por

tuga

l

Spa

in

Turk

ey

Pol

and

Phi

lippi

nes

Sou

th A

fric

a

Aus

tral

ia

Zam

bia

Bot

swan

a

Con

go D

.R.

US

A

Per

u

Nam

ibia

Arg

enti

na

Laos

Can

ada

Ind

ones

ia

Pap

ua N

ew G

uin

ea

Bol

ivia

Chi

na

Bul

garia

Mon

golia

Kaz

akhs

tan

Mor

occo

Bra

zil

Arm

enia

Chi

le

Vie

tnam

Sw

eden

Fin

land

Indi

a

Mau

ritan

ia

Uzb

ekis

tan

Mex

ico

Rus

sian

Fe

dera

tion

Iran

Sau

di A

rabi

a

Die

sel P

rice

(U

S$/

lite

r)

Diesel Price by Copper-Producing Location



Exhibit 151 Unsurprisingly, the Developed World Has the Highest Labor Cost; Cheap Labor (as Well as High Grades) Will Drive the Value of the African Copper Production; Chile Is Already a Medium- to High-Cost Mining Location

Source: ILO, BLS and Bernstein estimates and analysis.

Exhibit 152 While Diesel and Power Are Tied to Global Pricing Trends, Labor Shows the Highest Variation Across Regions

Source: ILO, BLS and Bernstein estimates and analysis.

0

1000

2000

3000

4000

5000

6000S

audi

Ara

bia

US

A

Sw

ede

n

Fin

lan

d

Ca

nada

Aus

tral

ia

Po

rtug

al

Spa

in

Sou

th A

fric

a

Turk

ey

Ira

n

Pol

and

Ru

ssia

n F

ede

ratio

n

Na

mib

ia

Arg

ent

ina

Ch

ile

Bo

tsw

ana

Per

u

Bra

zil

Ka

zak

hsta

n

Bul

garia

Mo

rocc

o

Ch

ina

Mex

ico

Ind

ones

ia

Arm

enia

Mon

golia

Bol

ivia

Pap

ua

Ne

w G

uin

ea

Mau

ritan

ia

Uzb

eki

stan

Indi

a

Phi

lippi

nes

Zam

bia

Laos

Vie

tnam

Co

ngo

D.R

.

Lab

or

Pri

ce (U

S$

/mo

nth

)

Labor Price by Copper-Producing Country

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Diesel Power Labor

Co

effi

cie

nt o

f V

aria

tio

n (s

td d

evi

ati

on

to m

ean

)

Global Variation in Main Drivers of Copper Mining Costs



What Is Embedded in Consensus? Before we explain how we forecast component cost escalators, it is worth drawing attention to the cost evolution anticipated by the most influential independent copper industry consultants. It is precisely these assumptions that will in turn be reflected in the consensus price expectations (see Exhibit 153 through Exhibit 156). We note that these are nominal escalators! Despite the fact that the world's financial system is driven by the expectation of nominal growth (as the talk surrounding the nominal GDP target after Mark Carney's appointment as governor of the Bank of England testified to), consensus anticipates that all input costs in copper mining will either fall or track flat from today to perpetuity. This assumption then translates into the declining consensus price expectation. In our view, this reasoning just commits the logical fallacy of begging rather than answering the question.

Exhibit 153 A Belief in Declining Labor Costs Is at Odds With the Reality of Demands for Increasing Nominal Wages

Exhibit 154 Power Costs Are Historically Highly Correlated With Oil Prices...


Exhibit 155 ...And a Belief in an Immediate and Sharp Downward Correction Is Not the View We Endorse

Exhibit 156 The Aggregate Cost Escalator That Stands Behind the Consensus View of the Copper Price Seems, at Least to Us, Highly Implausible


0

50

100

150

200

250

300

350

400

450

2000

2002

2004

2006

2008

2010

2012

2014

E

2016

E

2018

E

2020

E

2022

E

2024

E

1990

= 1

00

Wood Mackenzie Indexed Nominal Labor Escalator

0

50

100

150

200

250

300

350

400

450

2000

2002

2004

2006

2008

2010

2012

2014

E

2016

E

2018

E

2020

E

2022

E

2024

E

1990

= 1

00

Wood Mackenzie Indexed Nominal Power Escalator

0

100

200

300

400

500

600

700

200

0

200

2

200

4

200

6

200

8

201

0

201

2

20

14E

20

16E

20

18E

20

20E

20

22E

20

24E

199

0 =

10

0

Wood Mackenzie Indexed Nominal Diesel Escalator

050

100150200250300350400450

200

0

200

2

200

4

200

6

200

8

201

0

201

2

2014

E

2016

E

2018

E

2020

E

2022

E

2024

E

19

90 =

100

Wood Mackenzie Indexed Nominal Other Costs Escalator

In Contrast to Consensus, We Do Not Believe That All Input Costs in Copper Mining Will Either Fall or Track Flat from Today to Perpetuity



While it could be reasonable to expect that power and energy will suddenly become abundant as resource depletion and demand growth take their toll, this is certainly not the view of Bernstein (see Global Oil Prices: At "Base Camp" Before the Final Ascent). Still, we would concede that the path for energy costs is uncertain and that a case for declines in diesel and power prices could be made. However, it is on the issue of labor costs that an expectation of decline, in our view, is completely unfounded. In Exhibit 157, we show the history of nominal wages in the U.S. since 1900. An expectation of growth in wealth (at least in name, if not in reality) is critical to the political well-being of an industrial society. One has to look at only the demands from union representatives all over the world to see how ingrained this expectation is irrespective of the inherent cyclicality in mining. Consequently, we take the expectation of an upward trajectory in nominal wages denominated in the local currency as a fact of life. To us, the only questions are how these increases will vary from country to country, what their articulation will be in USD terms and, consequently, how they will impact the real copper price. In Exhibit 158, we show our forecasts for labor cost increases for the key mining jurisdictions and contrast them with the expectation that we believe informs the consensus copper price line.

Given that a multitude of low-grade and labor-intensive Chinese mines sit on the right-hand side of the global copper cost curve, the key determinant of its future marginal cost (and thus, price) will be the future evolution of these mines. Given the global nature of oil and energy price, a view on this evolution boils downs to a view on the future development of the Chinese mining labor costs. First, we expect the RMB to appreciate (see Exhibit 159), as the Chinese economy develops and the RMB becomes more firmly established as a store of value. Second, we also expect that the nominal local currency growth in the Chinese economy will continue at pace (see Exhibit 160). However, this growth will be supported by a declining workforce (see Exhibit 161). All these factors imply a very significant increase in the USD-denominated output per active Chinese worker (see Exhibit 162).

The only way for this not to lead to significant increases in wages is if all production growth accrues to capital rather than the Chinese labor (i.e., labor subsidizing capital). In our view, this is highly unlikely. First, the degree of subsidy is just too great for it to be realistic. Second, political implications of such a subsidy would be entirely unacceptable.

SCB's O&G Team Shares Our View on the Future Oil Prices; With Labor as a Key Driver Behind Mining Costs, We View an Expectation of Declining Labor Costs as Unreasonable

Exhibit 157 Nominal Wage Increases and the Expectation of a Higher Living Standard Over Time Form the Basis of the Political Mandate in Most Developed and Developing Countries

Source: BLS and Bernstein estimates and analysis.

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

1900

1905

1910

1915

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

Ind

exed

US

Wag

e R

ate

(190

0 =

100)

US Wages Over Time



Exhibit 158 We Take a Different View from Wood Mackenzie (and Hence from Consensus) on How Mining Costs Will Evolve, With the Differences Being Particularly Pronounced for Expected Future Labor Costs


Exhibit 159 Our Base Case Has Continued Appreciation in the RMB...

Exhibit 160 ...And Significant Growth in Nominal RMB-Denominated Output

Source: IHS and Bernstein estimates and analysis. Source: IHS and Bernstein estimates and analysis.

-4%

-2%

0%

2%

4%

6%

8%

10%

12%

14%

China Chile Peru USA Zambia Wood Mac

La

bo

r C

ost

CA

GR

2013-2025 Nominal Labor Cost Escalator

5.13

6.18

+20%

20252013

RMB Appreciation

195,024

58,568

20252013

+233%

Nominal RMB GDP (bn)



Mining Cost Escalation In our analysis, we assume that the labor share of output stays constant and that wages will rise in line with output per worker. One of the key arguments behind the slowing rate of capitalization in China (i.e., industrialization) is that the country is approaching the Lewis tipping point (Euro Metals & Mining: China's industrial behemoth with bones like iron bars - fuelled by inexpensive labor for how long?). This means that the historically high supply of cheap rural labor (which has acted to subsidize the returns to capital and support high rates of capital formation previously) is running out. As the cheap labor becomes ever scarcer, the rate of capital formation will slow down. However, the deceleration is possible only because of the increasing returns to labor. If remuneration did not rise, there would be no brake imposed on the rate of capital formation from declining labor supply. Consequently, the flip side of the deceleration in the Chinese fixed asset investment growth ought to be the acceleration in the share of output claimed by labor. Although we recognize it as a source of upside potential to our copper price forecast, to err on the side of caution, we have not assumed this in our analysis.

We repeat this labor-cost analysis for all copper-producing regions. We also add the SCB view on the global price of oil and energy (see Exhibit 163 and Exhibit 164) in proportion to the weight of each cost component in the overall cost structure of each step in the copper mining value chain. Subsequently, we calculate an escalator for each cost category and generate an overall copper mine cost escalator on a country-by-country basis (see Exhibit 165 through Exhibit 168).

Exhibit 161 At the Same Time, We Expect the Chinese Workforce to Continue to Decline

Exhibit 162 Overall, We Expect the Dollar Output per Chinese Worker to Increase Significantly, and This Drives Our Forecast for the Chinese Mining Labor Cost Escalation

Source: IHS and Bernstein estimates and analysis. Source: IHS and Bernstein estimates and analysis.

By Repeating the Labor-Cost Analysis for Each Copper-Producing Region and by Adding the SCB View on Oil and Energy Prices in Proportion to the Weight of Each Cost Component in the Overall Cost Structure, We Construct an Overall Copper Mine Cost Escalator on a Country-by -Country Basis

764790

‐3%

20252013

Work Force Decline (m)

49.70

12.70

+291%

20252013

US$ Output Per Worker



Exhibit 163 Added to This Is the SCB View on the Evolution of the Oil Price...

Exhibit 164 ...Which We Believe Will Imply a Continuation of the Increase in the Nominal Price of Energy


Exhibit 165 Adding These Factors Together Gives a Country-Specific Mining Cost Escalator...

Exhibit 166 ...As Well as Milling Cost...


0

100

200

300

400

500

600

700

800

900

199

0 =

10

0

Wood Mackenzie vs. SCB Indexed Nominal Diesel Escalator

SCB Wood Mac

0

100

200

300

400

500

600

199

0 =

100

Wood Mackenzie vs. SCB Indexed Nominal Power Escalator

SCB Wood Mac

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

7.0%

8.0%

Average Mining Cost Growth 2013-2025

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

Average Milling Cost Growth 2013-2025



Exhibit 167 ...And Ancillary Costs... Exhibit 168 ...For a Total Cost Escalator for Mined Copper on a Region-by-Region Basis


As we have stressed before, it is differential rather than absolute movements in the cost curve (i.e., rotations rather than translations) that generate margin. Furthermore, if high-cost growth is associated with low-cost regions and low-cost growth is associated with high-cost regions, the effects flatten the cost curve and reduce (rather than increase) the margin. In the first instance, we show how the copper mining costs in the U.S. compare to those in China (see Exhibit 169). In real terms, we expect the former to track roughly flat and the latter to rise dramatically. Given that the Chinese production already sits on the right-hand side of the global cost curve, this would lead to a steeper (rather than flatter) copper cost curve. Consequently, not only will the productivity of the global copper industry decline as the impact of deeper mines and lower grades makes itself felt, but the costs associated with the lower productivity will also rise (see Exhibit 170 through Exhibit 173). As a result, in our view, copper price must rise in real terms, thus generating further margin for the mining industry. The only way this can be avoided is if there is some kind of geological or technological step-change in the industry. Any geological step-change would simply take too long to make itself felt over the forecast period (the lead time between new discovery and first production is simply very long). In terms of technological step-change, we see nothing on the immediate horizon that would have the required potential (in situ acid leaching is unproven and deep sea mining is deemed as uneconomic and too small). Any even-handed analysis of the copper industry ought to conclude that nominal input costs will rise (rather than fall) and that there is nothing to offset the impact of declining mined grades. In such a situation, US$10,000/t becomes ever more inevitable.

0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

Average G&A Cost Growth 2013-2025

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

7.0%

8.0%

Total Copper Cost Growth 2013-2025

The Only Way Copper Price Increases Can Be Avoided Is if There Was Some Geological or Technological Step-Change in the Industry



Exhibit 169 It Is the Differential Cost Escalation That Drives Real Commodity Price Increases in USD Terms

Source: Wood Mackenzie, IHS and Bernstein estimates and analysis. Exhibit 170 The Increase in Cost Pressure Is Powerfully

Attested to by the Increase in Ore That Needs to Be Treated...

Exhibit 171 ...And the Declining Grade of the Ore That Is Processed


2.89%

6.95%

+140%

ChinaUSA

Differential Cost Escalation USA to China

1,000

1,500

2,000

2,500

3,000

3,500

2005

2006

2007

2008

2009

2010

2011

2012

201

3E

201

4E

201

5E

201

6E

201

7E

201

8E

201

9E

202

0E

Mt

Ore

Global Ore Milled

0.40%

0.50%

0.60%

0.70%

0.80%

0.90%

1.00%

2005

2006

2007

2008

2009

2010

2011

2012

201

3E

201

4E

201

5E

201

6E

201

7E

201

8E

201

9E

202

0E

Cu

Gra

de

(%)

Global Ore Grade



Exhibit 172 It Implies Lower Mill Recoveries Exhibit 173 All These Factors Feed Into Our Expectation of Decreasing Labor Productivity and a Macro Environment With Ever-Increasing Nominal Wages


Summary of Supply Side If we summarize the totality of the arguments presented so far looking at the supply side of the industry, the supply side story runs as follows. Geological degradation and declines in the quality of mined copper ore is, in our

view, inevitable. It will push the price of the metal up, unless this effect is offset by mining input cost declines.

Over time, we expect input costs to increase rather than fall, driven by nominal wage increases as well as continued energy scarcity.

Both expected cost escalation and cost position make us anticipate a steeper rather than flatter future global copper cost curve. The high-cost Chinese mines are the most exposed to cost escalation and also sit at the high end of the cost curve.

All these factors suggest that copper prices will have to rise in real terms and offer compelling support to the thesis that the world will soon face US$10,000/t copper price. However, the final piece of the puzzle (on the supply side at least) requires an analysis of the incentive price structure of the industry. The threat is that as prices rise, a new "wave" of supply from projects that are already waiting in the wings will hit the market, therefore postponing the date that will see copper at US$10,000/t. Our analysis has already demonstrated that new copper supply is insufficient to derail the eventual realization of US$10,000/t, as there is simply not enough geological endowment that could enable this. Chile is unique, and no country will replicate the impact it had on the global copper market. However, it is possible that a wave on new supply could postpone the day of reckoning. It is to this risk that we turn our attention in the final piece of our supply side analysis.

83.0%

83.5%

84.0%

84.5%

85.0%

85.5%

86.0%

86.5%

87.0%

87.5%20

05

2006

2007

2008

2009

2010

2011

2012

201

3E

201

4E

201

5E

201

6E

201

7E

201

8E

201

9E

202

0E

Cu

Rec

ove

ry (

%)

Milled Ore Recovery

2.5

2.7

2.9

3.1

3.3

3.5

3.7

3.9

4.1

2005

2006

2007

2008

2009

2010

2011

2012

201

3E

201

4E

201

5E

201

6E

201

7E

201

8E

201

9E

202

0E

To

ns

Mil

led

pe

r M

an H

ou

r

Global Copper Productivity



Structure of the Incentive Price Curve — What Is Needed to Clear the Market Bob Dylan once sang that "what looks large from a distance, close up ain't never that big." These words definitely relate to copper projects. Far-dated supply potential always looks impressive, but the closer you get to it, the less realistic it becomes. Aynak is the latest example of a project that demonstrates why today's price environment is insufficient to close any future supply/demand imbalance.

We would like to simply recap what we have briefly touched upon earlier regarding current copper prices not being high enough to incentivize new capacity to come on line. We examine the cash and capital costs of approximately 300 greenfield projects and the associated brownfield expansions. Unsurprisingly, there is significant operating margin to be made through capacity expansion. However, we also note that the capital requirement is such that against a fully loaded discount rate (i.e., including an adjustment for country risk), it is only at prices above US$8,000/t to US$9,000/t that the majority of them will be value accretive (see Exhibit 174). Today, more than ever, shareholders are aware of the risks inherent in new greenfield projects (e.g., Pascua Lama, Minas Rio and Riversdale). Consequently, many investors have been demanding that capital is returned rather than expended in an effort to push commodity prices down. While we have not factored these strategic issues into our analysis of the returns required for new investment (basing them on our understanding of the investment protocols in the large mining houses), they support our belief that new copper project approvals will struggle. Moreover, in a period of declining or flattening commodity prices, concerns of value dominate the thoughts of the miners ahead of growth. This acts as a break on new volume expansion.

Only at Prices Above US$8,000/t to US$9,000/t Will the Majority of Projects Be Value Accretive; Given the Current Price Environment, Miners' Reluctance to Approve New Projects Will Persist Until Copper Prices Go Up

Exhibit 174 Against a Fully Loaded Cost of Capital, It Is Clear That Many of Today's New Projects Will Struggle to Create Value


10,0000 2,000 6,0004,000 8,000 12,000 14,000 16,000 18,000 22,000 24,000

17,000

16,000

15,000

14,000

13,000

12,000

11,000

10,000

9,000

6,000

7,000

8,000

5,000

4,000

3,000

2,000

1,000

0

20,000


Incentive Price –US$/t



However, there is more than just the aggregate structure of the copper incentive price curve that needs to be considered. The incentive price curve gives a snapshot of the entirety of all known and possible sources of supply. However, the evolution of price is necessarily a temporal sequence, yet there is no temporal component in the incentive price curve. There is also a financial barrier to investment in copper projects, which is what the incentive price does capture — and there is a technical barrier. Not all projects are equally well developed — some are at a very early stage and are not investment-ready, others are already in possession of a bankable feasibility study. It could well be the case that while the average incentive price is high, the projects that are investment-ready and well advanced have a low incentive price while those that are far-dated options have a high incentive price. It would be exactly this circumstance that could lead to a situation where rising prices trigger a wave of new supply that (at least temporarily) halts the rise in copper prices.

Nevertheless, this is not what we observe. In fact, the distribution of incentive prices over time is remarkably constant (see Exhibit 175). Therefore, while post 2016 there is sufficient copper to cover demand, it is sufficient only at prices that afford developers a reasonable return. It is some way higher than the price today.

We regard the structure of the incentive price curve as generating a bare minimum long-term price for two further reasons. The natural tendency for project costs to rise, as they move through the stage gate

process from conceptual to prefeasibility and to feasibility. Not all conceptual projects pass the screening of a prefeasibility study, and not all prefeasibility study projects make it to a feasibility study. This tells us that the economics of mining projects have a tendency to deteriorate (rather than improve) as the project moves toward technical viability. Consequently, the further dated options that form part of the set of projects used to calculate the incentive price probably underestimate (rather than overestimate) their true incentive price.

The cost structure in Exhibit 174 is denominated in today's money. However, capital costs (as well as operating costs) are subject to cost escalation. A very significant part of the overall capital cost of a project is the capitalization of labor

We Observe That the Distribution of Incentive Prices Over Time Is Remarkably Constant, Hence Not Allowing for an Unexpected Wave of Incremental Supply to Halt Copper Price Escalation

Exhibit 175 There Is Always a Bias to Execute the Best Projects First; Even So, a Real Price of Above US$9,000/t Will Be Required to Bring the New Projects On Line Over the Next Decade


Incentive Price Curve Is an Indicator of ''the Bare Minimum Long-Term Price'' Required to Incentivize Further Capital Investment

0

500

1000

1500

2000

2500

3000

3500

0

2,000

4,000

6,000

8,000

10,000

12,000

2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024

Max

imu

m A

vai

lab

le N

ew

Ca

pac

ity

(kt

)

Ince

nti

ve P

ric

e o

f N

ew

Ca

pa

city

(US

$/t)

Distribution of Possible New Projects by Start Date

Max New Volume Average Incentive Price



that has been expended in construction. Just as in the analysis of operating costs presented previously, we ought to expect real-term labor cost increases to raise capital intensities. Moreover, capital costs (just like operating costs) are geared to grade. For example, the milling circuits scale with the tons milled, yet metal output is a product of tons milled and ore grade.

A key part in any incentive price analysis is, of course, the discount rate used to establish the price of metal that yields a zero NPV. While country risk (strictly speaking) should be handled as a cash-flow adjustment (and should be valued through the price of political insurance or the cost of greater security), it is common practice to compensate for it through an adjustment to the discount rate, which is the practice we adopt here. When decisions are made on project investments, very few will attract investment if they just meet the cost of capital. Instead, they must exceed it and also exceed it after adjustment for country risk. Exhibit 176 shows the country risk premiums that we use in constructing our incentive price curve. There is (in general) a negative relationship between the country risk and its copper production (see Exhibit 177). This tells us that mining investment has preferentially targeted those locations that are the easiest to manage from a political perspective. In addition, it shows that our expectation of how easily and rapidly new projects can be brought on line is likely to be highly influenced by the experiences of the recent past. The last few decades of copper growth have been relatively easy, but only because of the bias towards investment in easy locations.

Due to the Way Project Investment Decisions Are Made, Many Find Themselves Overly Optimistic About How Easily New Projects Can Be Brought On Line Given That All of the ''Easy'' Locations Have Been Highly Exploited

Exhibit 176 A Key Component of the Required Price Is Due to the Requirement to Compensate for the Risk and Losses That Will Be Borne as a Consequence of More Supply Coming from New Frontier Regions

Source: Frasier Institute and Bernstein estimates and analysis.

0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

14.0%

Ven

ezue

laA

fgha

nis

tan

DR

CR

us

sia

n F

ede

rati

onA

rme

nia

Aze

rbai

jan

CIS

Ira

nM

yan

ma

rP

akis

tan

Phi

lippi

nes

Sou

th A

fric

aK

azak

hst

anK

yrg

yzs

tan

Mo

ngol

iaIn

dia

Ind

ones

iaB

otsw

ana

Erit

rea

Na

mib

iaS

audi

Ara

bia

Zam

bia

Pap

ua

Ne

w G

uine

aC

hin

aA

rgen

tina

US

AV

ietn

amD

om

inic

an R

epub

licE

cuad

orM

exi

coP

anam

aS

erbi

aA

ust

ralia Fiji

Gre

ece

Isra

elP

ortu

gal

Rom

ania

Spa

inU

nite

d K

ingd

omP

eru

Ca

nad

aB

razi

lS

wed

enC

hile

Fin

land

Dis

co

un

t R

ate

Ad

just

men

t

Country Risk Premium by Copper-Producing Region



The very recent example of Aynak just shows how difficult new copper development will be in higher-risk locations. This Afghan deposit, discovered in the 1970s by Soviet geologists on the site of ancient copper workings, was acquired by Chinese investors (Metallurgical Corp of China and Jiangxi copper) in 2007. The concession was won amid allegations of impropriety after an agreement was reached that would see the winning consortium invest US$3.7 billion to develop the project. Putting aside the requirement to demolish the archaeological remains of Afghanistan's oldest Buddhist monastery to make way for the mine, the consortium also agreed to construct rail infrastructure, a power plant and a copper smelter as part of the development, not to mention the 20% royalty. In 2009, the expectation was that commercial copper production would start in 2011. In 2012, this was revised to a start date of year 2014. The latest publicly available capital cost estimate has seen costs rise to US$4.4 billion. This month, the consortium has demanded a review of the entire deal and has asked that the royalty rate be halved to 10%, the requirement to build the smelter and power plant be cancelled, and the requirement to lay the railway line postponed. Furthermore, production from the mine is delayed to 2011.4 It is hard to speculate on exactly what the latest capital

4 http://www.miningweekly.com/article/landmark-chinese-copper-deal-with-afghanistan-at-risk-2013-08-27.

Exhibit 177 The Current World Production of Copper Has Been Achieved Through Targeting the Easy Win Locations, Which Combined High Geological Prospectivity With Low Political Risk; This Model No Longer Holds and Higher Prices Will Be Required to Reflect This Reality


Aynak Proves to Be a Good Example of the Above Described Phenomenon

1,400 1,500 1,600 5,900200

2.5

3.5

4.0

4.5

5.0

5.5

6.0

1.5

3.0

2.0

8.0

7.5

7.0

6.5

8.5

9.0

9.5

10.0

1,3001,1001,000900800700600500400300 1,2001000

1.0

0.5

0.0

Argentina

Mongolia

USA

Canada

Country Risk Premium ‐ %

Peru

Chile

Kazakhstan

Iran

Afghanistan

CIS

Ecuador

Brazil

Panama

Zambia

Mexico

Philippines

Existing Copper Production

Australia

DRC

Papua New Guinea

China

Diameter Represents New Project Capacity



costs estimates are given the change of events, but the requirement to cancel all associated project infrastructure tells us everything that we need to know. At today's copper price even an operation seeking to exploit 1.6% Cu grade in an area where labor will essentially be free is incapable of meeting the cost of capital of a Chinese investment consortium (whatever that may be).

If there is no wave of new capacity in the medium term, the final threat to higher copper prices (given a normalized demand environment) is removed. The cost curve will steepen and prices will rise, which is exactly what stands behind our US$10,000/t copper prediction (see Exhibit 178). Clearly, all of this is predicated on the return of a "normalized" demand environment.



Exhibit 178 We Believe That Low-Grade High-Cost Chinese Producers Already Sit at the Right-Hand Side of the Global Cost Curve and That Their Presence There Is Poorly Understood; As Real USD Cost Escalation in China on the Back of Increasing Labor Costs Takes Hold, It Will Steepen the Cost Curve and Lead to Ever Higher Copper Prices


-

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

20,000

0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 55% 60% 65% 70% 75% 80% 85% 90% 95%100%

Cas

h C

ost

s C

1+S

IB C

apex

-U

S$/

t -N

om

inal

Cumulative Mined Production as % of Demand

2012 2017 2022



Demand — Waiting for the Trend to Reassert Itself

We believe that copper is a "later cycle" commodity than steel. Consequently, we have a more positive view on the future demand for copper compared to iron ore. There are more demand drivers in copper than in iron ore, which is heavily skewed to the two sectors of infrastructure and automotive demand. To arrive at our view on demand growth, we continue to look at more than just the copper consumption rate. Critically, we examine the total amount of metal embedded in the capital stock of a country. For China, we calculate copper stock of 29kg/capita, which is equivalent to just 30% of Japanese levels and 22% of South Korean levels. Moreover, whatever is taking place in China today, it is not the same as Japan's "lost decade" in 1990s. By 1990, Japan had completed the installation of its capital stock (copper stock stood at 92kg/capita compared to 94kg/capita today). Knowing the end to which copper consumption is oriented — namely, the development of an industrial society — enables us to forecast copper intensity evolution based on something more scientific than simply the aesthetics of curve drawing.

What Does History Tell Us? On the demand side of the copper market analysis, the main concern has been the relatively muted volume growth seen over the last three years (1.2% CAGR since 2010). This immediately raises the question of whether we think that 2013 represents either: The point at which copper goes ex-growth globally? Or the point at which there is a downward step-change in demand growth relative

to the last decade? The answer to the first question is an emphatic "no." Unlike steel, we have

never seen a period in which global copper demand goes ex-growth on a sustained basis (see Exhibit 179). The belief that the last three years have been something more profound than the periodic volatility affecting all commodity markets seems to us, at least statistically speaking, highly unconvincing. However, the second question is by far more interesting and we dedicated this chapter to it.

To begin with, trajectories taken by copper and steel consumption post the oil shocks and following the rise of China have been markedly different. We divide this into four very broad periods (see Exhibit 180 through Exhibit 182): Period 1 (1900 to 1948) — Initial capital build in the West. Period 2 (1948 to c. 1975) — Post-war reconstruction of Europe and integration

of Japan. Period 3 (1975 to c. 1995) — End of Western industrialization and change in

economic model precipitated by oil shocks. Period 4 (1995 to present) — The start of Chinese industrialization.

For steel, the punctuation in demand in period 3 is much more severe than for copper. Following this, as we enter period 4, the trend demand rate for steel returns to almost exactly the same trajectory as before the oil shocks/Cultural Revolution (see Exhibit 183).

However, copper shows a markedly lower demand growth than might have otherwise been expected. Prior to the start of China's industrialization, steel and copper growth rates had tracked each other very closely. Since the turn of the century, copper demand growth has moved from being at parity with steel to representing only a fraction of steel growth. Consequently, we believe that copper is genuinely a "later cycle" commodity whose growth is predicated on a wider base

Demand for Copper Will Be Well-Supported in the Medium Term

Copper Is a ''Later Cycle'' Commodity Whose Growth Is Predicated on a Wider Base of Applications Than Steel



of applications than steel. While it may not enjoy the same growth rate as steel in the initial phases of industrialization, it will not be so quickly displaced once the economy begins to move beyond the initial capital accumulation phase.

Exhibit 179 Post the Oil Shocks of the 1970s, There Was a Step-Change in Demand for Copper; While Growth Slowed, It Never Went Ex-Growth in a Trend Sense


Exhibit 180 Global Copper Growth Is Still Not as Strong as It Was Post World War II...

Exhibit 181 ...With the West Acting as a Drag on Overall Demand



2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

20,000

22,000

190

0

190

5

191

0

191

5

192

0

192

5

193

0

193

5

194

0

194

5

195

0

195

5

196

0

196

5

197

0

197

5

198

0

198

5

199

0

199

5

200

0

200

5

201

0

Re

fin

ed C

op

per

Dem

an

d (

kt C

u)

Global Refined Copper Demand

Actual Refined Cu Demand Trend Cu Demand Continuation of Pre-1975 Trend

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

1900-1948 1948-1972 1972-1991 1991-2012

Global Copper Demand Growth CAGR

-2.0%

-1.0%

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

1900-1948 1948-1972 1972-1991 1991-2012

USA, Canada, Japan and Western Europe Refined Copper Demand Growth CAGR



Exhibit 182 China and Developing Asia Have Taken Over the Lead in Copper Demand

Exhibit 183 The Current Cycle Shows Markedly Lower Copper Growth Than Steel



First Versus Final Copper Use One of the problems with any commodity demand analysis relates to disentangling first versus final demand. The classic example of this is that the copper in an air conditioner built in China, but exported to the U.S. will appear as "first use" demand for China but is really "final use" demand for the U.S. (see Exhibit 184). In general, the lower the value of the commodity relative to global freight rates, the less of an issue "first versus final use" becomes. Hence, while not critical for iron ore, we have to account for this in copper demand estimation.

We could do this by tracking inflows and outflows (possibly through multiple cycles of entry and exit of the same material) of all metal-bearing goods. However, we believe that this task is unfeasibly complicated, given the errors associated with each of the inevitable assumptions. Attempted in detail, this task usually ends with so large a degree of uncertainty that it renders the attempt moot. Accordingly, we adopted a method for adjustment that, although simple, we believe captures the essence of the situation. We divide copper into "exportable" and "non-exportable" forms. This comes

down to estimating metal used in "goods" versus that used in infrastructure and buildings (see Exhibit 185).

We assume that the overall level of economic activity is homogeneous (i.e., that copper in "goods" is as likely to arise in the part of GDP based on exports as in the part based on domestic consumption). Although this assumption is debatable for China, we believe it is first order correct (see Exhibit 186). Nevertheless, we hope to explore this issue in more detail in subsequent research.

For each exporting country, we calculate the size of the available export market based on the proportion of global trade (i.e., what does each export destination look like when viewed through the eyes of a Chinese exporter?). Hence, for each country, we calculate the degree of net copper imports and exports (see Exhibit 187).

We arrive at the estimated "final use" copper demand history for each location analyzed in our demand forecast.

While China is clearly a massive exporter of goods containing copper, its largest import category is electrical machinery (US$200 billion p.a.). Metal ores represent one-third on the list of imports (US$85 billion p.a.).

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

30.0%

1900-1948 1948-1972 1972-1991 1991-2012

China Refined Copper Demand Growth CAGR

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

1900-1944 1944-1972 1972-1999 1999-2010

Steel Versus Copper CAGRs

Copper Steel

The Lower the Value of the Commodity Relative to Global Freight Rates, the Less of an Issue the "First Versus Final Use" Becomes; Thus, While Not as Crucial for Iron Ore, We Have to Account for It in Copper Demand Estimation



Exhibit 184 How Much of the First Use Copper Demand Decline Is Attributable to Imports of Metal in Other Forms?

Exhibit 185 We Believe That Only About 50% of Copper Use Is "Exportable"

Source: Wood Mackenzie, USGS and Bernstein analysis. Source: WCGS and Bernstein analysis.

Exhibit 186 In Our View, Net Export to Import Positions Are Representative of the Overall Flow of Copper in Goods

Exhibit 187 This Yields the Following "Finished to Final" Use Correction

Source: WCGS and Bernstein analysis. Source: WCGS and Bernstein analysis.

0200400600800

1,0001,2001,4001,6001,8002,0002,2002,4002,6002,8003,0003,200

kt

US First Use Refined Copper Consumption

33%

33%

13%

13%

8%

Global Copper Consumption by End Use

Construction Electrical Applications

Industrial Machinery Transportation

Consumer Products

0%

5%

10%

15%

20%

25%

30%

35%

USA China

Export/Import Position, China and US

Exports % of GDP (2011) Imports % of GDP (2011)

-800

-600

-400

-200

0

200

400

600

kt

Net Cu Imports/(Exports) in Goods



Copper Consumption, Capital Stock and the Derivation of Intensity Curves We look at two metrics to understand the current stage of a country's copper industry: Current copper intensity: We define this as copper consumed per unit of output

(i.e., copper per US$1,000 GDP), against the overall level of output (i.e., GDP per capita). We think of it as a "rate" term, showing how rapidly the country is embedding copper into the economy in a given period.

Cumulative copper intensity or copper capital stock. We define this as the total historical copper installed per capita against the overall level of output (again GDP per capita). We look at this as a measure of the "level" of the economy's development, looking at how successful historical capital investment decisions have been (i.e., the current level of output being generated by historical investment decisions).

We believe that both of these measures are necessary to gauge the current position of copper (and other commodities') consumption in an economy, as the "rate" term by itself only tells how fast an economy is developing. However, it does not tell us whether this development is to be regarded as successful or for how long that rate can be sustained. Only by looking at the "level" term can we answer the question about the success and longevity of the industrialization. We also derive a third measure of copper consumption per capita. It is best thought of as a derivative relationship rather than a causative one, as the basis of its relationship to economic activity is unclear. Put simply, we believe that it is impossible to use the history of copper intensity alone (whether measured per unit output or per capita) to derive a forecast for future copper intensity. There is simply not enough information in the "rate" term of copper consumption to solve the problem of where any trajectory is ultimately heading. Another piece of information is required. We find that information in the role played by copper stock in supporting the generation of wealth.

With that in mind, we take China's Asian neighbor Japan as an exemplar of how to derive the relationship between copper demand growth and overall economic growth (we could have also chosen the U.S., but given the change in the role of copper use brought by electrification, we wanted a parallel closer in time). Exhibit 188 shows the history of Japanese refined copper consumption since 1900 and Exhibit 189 converts that data from a temporal to an economic sequence. This data shows the paradigmatic path that we see in all analysis of metal intensity through the course of economic development. It can be divided into three periods: Development — During the build up of industrialization, copper use grows faster

than underlying GDP. This corresponds to the installation of a capital base that will subsequently generate output.

Peak — The "peak" refers to a peak in intensity, where GDP growth and copper growth are matched. During this phase, we see the effect of diminishing returns — additional capital spending starts to deliver incrementally less output.

Decline — The development of an economy is based on tertiary forms of value-add or forms of manufacturing that require little incremental consumption of resources. Once this occurs, overall economic growth will be faster than growth in metal consumption.

When Understanding a Country's Copper Industry, We Think About the Problem in Terms of Current Copper Intensity and Cumulative Copper Intensity; the Latter, We Believe, Is Often Ignored by Many

The Relationship Between Copper Demand Growth and Overall Economic Growth Can Be Divided Into Three Periods: Development, Peak and Decline



Exhibit 188 Japan Presents an Interesting Precedent for Current Chinese Copper Consumption

Source: Wood Mackenzie, USGS, Mitchell and Bernstein estimates and analysis.

Exhibit 189 The Copper Consumption in Japan Can Best Be Understood by Looking at Copper Intensity


More than just the rate of copper consumption, we can look at the total amount of metal embedded in the capital stock of Japan by taking the cumulative metal consumption after adjusting for losses through depreciation (see Exhibit 190). The productivity of the economy starts increasing after a certain critical level of capital stock is reached (~100kg per capita). The current levels of Japanese copper consumption keep track with depreciation, but do not materially add to the total capital stock. Instead, economic growth occurs as a function of utilizing the

-

200

400

600

800

1,000

1,200

1,400

1,600

1,800

1900

1905

1910

1915

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

kt C

uJapan Final Use Refined Copper

-

.10

.20

.30

.40

.50

.60

.70

.80

.90

- 5.0 10.0 15.0 20.0 25.0 30.0 35.0

kg

/'000

$ G

DP


Japan Copper Intensity

Japan Actual Japan Trend

Productivity of an Economy Starts to Increase After a Certain Critical Level of Capital Stock Is Reached (~100kg/capita)



installed capital more efficiently. This relationship enables one to assert the "end" to which copper consumption is oriented — namely, the development of an industrial. Unsurprisingly, we see a very strong relationship exists between the overall level of capital stock in a country and its output (see Exhibit 191). Moreover, it is only relatively late in a country's economic development that the bifurcation between a service and a manufacturing-oriented economy actually takes place. There is a certain base load of metal that must be installed before this economic "choice" is made. In this context, it is important to note that China has a capital stock of copper equivalent to 29kg/capita — just 30% of Japanese levels and 22% of South Korean levels. Whatever is taking place in China today, it cannot be the Japanese "lost decade" post-1990. By 1990, Japan had fully installed its capital stock (copper stock stood at 92kg/capita versus 94kg/capita today). China is nowhere near these levels today. We are not saying that its growth is not slowing. Rather, we aim to highlight the dangers of too simplistic a historical comparison between economies and political systems at different points in their development.

Knowing the end role of copper consumption provides the "missing" data point that is not supplied by an analysis of copper consumption rate alone. Consequently, the trajectory from the present into the future in terms of copper intensity can be constructed for any country. Clearly, the most important country for which we attempt this is China. In Exhibit 192 through Exhibit 194, we show the history of copper consumption, copper intensity and copper stock (respectively) from 1900 to today. The resultant pattern is all deeply familiar. The critical question is how to draw the path of forward-looking copper intensity shown in Exhibit 193. Apart from the aesthetics of line drawing, what makes any one trajectory for China more plausible than another? It is always possible to attempt to build a "bottom up" demand model and try to disguise the nature of the macro call being made along the lines of "if I know the number of washing machines made in China each year and the contained copper in each machine, I can calculate the copper contained in that demand leg." However, how does one project the figure of washing machine demand without reference to an underlying growth assumption? In our view, such calls substantially increase the forecast error and do not actually alter the central demand-side call — namely, that of Chinese growth. With that in mind, we use the canonical relationship between copper stock and output presented by the Japanese economy to model a demand-side trajectory for China. If we assume that copper productivity in China mirrors that seen in Japan, the development of China's copper stock and its economy in total ought to follow the pattern set out in Exhibit 195. This translates into the pattern for copper intensity development seen in Exhibit 196. The resultant China "final use" refined copper demand forecast is shown in Exhibit 197. We repeat this methodology for each of the forecast countries and then aggregate the result to get to our view on total global copper intensity and copper demand (see Exhibit 198 and Exhibit 199). While there is more to copper demand than just China (in contrast to demand for iron ore), it is nonetheless true that China is the most important element in global demand by some margin.



Exhibit 190 However, There Needs to Be a Way of Bridging Between Metal Intensity in One Time Period to Another; We Find That Bridge in Metal (Capital) Stock Formation


R² = 0.9945

0

20

40

60

80

100

120

- 5.0 10.0 15.0 20.0 25.0 30.0 35.0

Cu

Ca

pit

al S

toc

k -k

g/C

apit

a


Japan Copper Stock



Exhibit 191 Unsurprisingly, a Very Strong Relationship Exists Between the Overall Level of Capital Stock in a Country and Its Output; We Link the Metal Intensity in Different Time Periods Through the Assumption That the Rate of Overall Economic Development Must Be Accompanied by a Corresponding Development in Capital Stock


R² = 0.6442

-

50

100

150

200

250

- 5 10 15 20 25 30 35 40 45 50

Co

pp

er C

apit

al S

tock

-kg

/Cap

ita

Output - GDP/Capital (Real 2005$ PPP)

Capital Stock of Copper to Overall Economic Output

South

Korea

USA

Japan

Germany

China



Exhibit 192 In Order to Forecast China's (as Well as Any Other Country's) Consumption, We Begin With the History of Copper Consumption...


Exhibit 193 ...And the Corresponding Copper Intensity...


-

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

1900

1905

1910

1915

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

kt C

uChina Final Use Refined Copper

-

.10

.20

.30

.40

.50

.60

.70

.80

- 5.0 10.0 15.0 20.0 25.0 30.0 35.0

kg

/'000

$ G

DP


China Copper Intensity

China Actual China Trend



Exhibit 194 …That Gives Rise to a Corresponding Development in Capital Stock


Exhibit 195 We Expect the Chinese Copper Stock Development to Resemble That of Japan; We Use This Expectation to Calculate Both How Copper Intensity in Any Developing Country Tracks the Overall Economic Development and the Multiplier Between Metal Growth and Economic Growth at Any Point in Time


R² = 0.9943

0

5

10

15

20

25

30

35

- 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0

Cu

Ca

pit

al S

toc

k -

kg/C

ap

ita


China Copper Stock

R² = 0.9945

0

20

40

60

80

100

120

- 5.0 10.0 15.0 20.0 25.0 30.0 35.0

Cu

Cap

ita

l Sto

ck -

kg

/Cap

ita


China Copper Stock Forecast

Japan Actual China Actual China Forecast



Exhibit 196 This Gives Rise to the Following Trend Line for China's Copper Intensity...


Exhibit 197 ...And a Corresponding Development in Its Copper Consumption; We Expect It to Peak at 14Mtpa Post-2025; in This Regard, Copper Is a "Later Cycle" Commodity Than Steel, Whose Peak We Anticipate Nearer 2020


-

.10

.20

.30

.40

.50

.60

.70

.80

- 5.0 10.0 15.0 20.0 25.0 30.0 35.0

kg

/'000

$ G

DP


China Copper Intensity Forecast

China Actual China Trend China Forecast

-

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

1900

1905

1910

1915

1920

1925

1930

1935

1940

1945

1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

kt C

u

China Final Use Refined Copper Forecast



Exhibit 198 We Then Aggregate Each of the Country-Specific Demand Forecasts to Arrive at a Global Total That Enables Us to Derive a Picture of Still Rising Copper Intensity Out Till 2020


Exhibit 199 Nonetheless, China Remains the Most Important Driver of Overall Demand


(0.5)

-

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

1910

1920

1930

1940

1950

1960

1970

1980

1990

2000

2010

2020

2030

10

Yr.

Avg

. C

op

pe

r g

row

th to

GD

P g

row

th

Copper Growth vs. Economic Growth

0

5,000

10,000

15,000

20,000

25,000

30,000

35,000

1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 2020 2030

Ref

ine

d C

op

pe

r -k

t Cu

Copper Consumption Forecast

Africa China India Japan Other Asia Europe Lat Am Middle East North Am Oceania



Price Forecast

We base our price forecast on the interplay between the cash cost curve, the incentive price curve, and the differential cost escalators that prevail between one mining jurisdiction and another. We see no reason to believe that the price agnosticism of consensus will hold true. In our view, copper will test the US$10,000/t mark by 2018.

We derive our price forecast through an analysis of the cost structure of the industry at each point in time. We identify the price that will be needed to clear the cost of the marginal unit of supply required to satisfy demand. We use the incentive price not as a measure of long-term price, but as the mechanism that determines the evolution of the shape of the future cost curve. The key components of our methodology for the copper price forecast are the following: We calculate operating cost curves for each year between 2005 and 2040 for 316

greenfield projects, 530 existing mines and associated brownfield expansions, the 53 “known” Chinese mines alongside estimates for structure of the "unknown" residual Chinese mines.

We look at the economically "rational" development of new projects based on an analysis of the earliest possible technical start date for that project and a comparison of the incentive price of that project with the prevailing price environment. We have the option of overlaying varying degrees of industry capital discipline by limiting the quantum of capital that the industry will expend in any given year (allocated to the highest-quality projects first), but in the base case, we leave this unconstrained.

We estimate country-specific real mining cost escalators through an explicit forecast of mining labor productivity, labor costs, consumable and power costs as well as currency appreciation (using Global Insight forecasts).

We piece together the future global cost curve by looking at the current structure of supply, the real cost escalation of each existing mine site and the modification to the cost curve that emerges as a consequence of the commissioning of new projects and the depletion of old assets.

We look at the price required to clear demand from all currently operating assets. The difference between this approach and that employed in our forecast for

iron ore (and other bulk commodities) lies in the analysis of trade flows. Given that freight rates are smaller than the value of copper, there is no real difference between CIF and FOB prices. Hence, we do not look at the impact that trade flows have on the price.

In order to calculate the relative competitiveness of each supply location (and hence the real cost escalator to be applied), we first break down the overall mining costs into their underlying components. This is shown for the mine site in Exhibit 200 and for the milling in Exhibit 201. As expected, labor, diesel and consumables (e.g., tires, explosives) form the largest cost element at the mine site, whereas electrical power, grinding media and reagents form the largest component of the milling cost base. Each of these elements is then separately forecasted to arrive at the overall cost escalator for each mine site in a particular geography (see Exhibit 202). The two important sources of differentiation between geographies arise from labor costs and any real currency appreciation or depreciation. Taken together, these escalators imply an overall cost escalator for each geography (see Exhibit 203). Unsurprisingly, Chinese costs show the greatest pressure on the back of the early stage of economic development and the rapidity of its growth.

Our Analysis Suggests That Copper Will Test the US$10,000/t by 2018



Exhibit 200 Labor (Including Services) and Consumables (Tires, etc.) Represent the Largest Cost Elements for the Mine Site...

Exhibit 201 ...While It Is Consumables (Grinding Media and Reagents) and Power That Form the Largest Cost Element for Milling

Source: Wood Mackenzie and Bernstein analysis Source: Wood Mackenzie and Bernstein analysis

Exhibit 202 The Variation in Labor Costs Is the Most Important Source of Differentiation for Cost Escalators

Exhibit 203 Adding in Currency Effects Generates the Overall Cost Escalator for Each Region

Source: Global Insight and Bernstein analysis. Source: Global Insight and Bernstein analysis.

26%

21%

20%

5%

28%

Mine Cost Composition

Labor Services Diesel Power Consumables

12%

16%

2%

32%

38%

Mill Cost Composition

Labor Services Diesel Power Consumables

0

50

100

150

200

250

300

350

201

2

2013

E

2014

E

2015

E

2016

E

2017

E

2018

E

2019

E

2020

E

2012

= 1

00

Operating Cost Escalators by Category

Steel Oil Labor - Chile

Labor - China Labor - USA

0

20

40

60

80

100

120

140

160

180

200

2012

201

3E

201

4E

201

5E

201

6E

201

7E

201

8E

201

9E

202

0E

2012

= 1

00

Operating Cost Escalators by Geography

Chile China USA



However, a consideration of cost escalators alone is insufficient to generate the future cost projection. Labor productivity must also be factored in (as with increasing labor productivity, more labor can be removed for the same level of output, thus obviating some of the impacts of labor inflation). Unfortunately, we believe that global mining productivity will continue to decline. We see no reason to believe that the underlying geological deterioration occasioned by exploitation of copper bodies should axiomatically be offset by increases in mining technology. Indeed, we believe quite the opposite — depletion and degradation are irreversible and represent a necessary component of ore extraction. On the other hand, technological improvement in mining has been achieved through ever more massive capitalization of mine sites. This factor is up against the inexorability of declining returns on capital. The key elements of this argument are laid out in Exhibit 204 through Exhibit 207.

Exhibit 204 The Increase in Cost Pressure Is Powerfully Attested to by the Increase in Ore That Needs to Be Treated...

Exhibit 205 ...And the Declining Grade of the Ore That Is Processed


Exhibit 206 It Implies Lower Mill Recoveries Exhibit 207 All These Factors Feed Into Our Expectation of Decreasing Labor Productivity and a Macro Environment With Ever Increasing Nominal Wages


1000

1500

2000

2500

3000

3500

200

5

200

6

200

7

200

8

200

9

201

0

201

1

201

2

201

3E

201

4E

201

5E

201

6E

201

7E

201

8E

201

9E

202

0E

Mt

Ore

Global Ore Milled

0.40%

0.50%

0.60%

0.70%

0.80%

0.90%

1.00%

20

05

20

06

20

07

20

08

20

09

20

10

20

11

20

12

201

3E

201

4E

201

5E

201

6E

201

7E

201

8E

201

9E

202

0E

Cu

Gra

de

-%

Global Ore Grade

83.0%

83.5%

84.0%

84.5%

85.0%

85.5%

86.0%

86.5%

87.0%

87.5%

2005

2006

2007

2008

2009

2010

2011

2012

2013

E

2014

E

2015

E

2016

E

2017

E

2018

E

2019

E

2020

E

Cu

rec

ove

ry -

%

Milled Ore Recovery

2.5

2.7

2.9

3.1

3.3

3.5

3.7

3.9

4.1

2005

2006

2007

2008

2009

2010

2011

2012

2013

E

2014

E

2015

E

2016

E

2017

E

2018

E

2019

E

2020

E

To

nn

es M

ille

d p

er

Man

Ho

ur

Global Copper Productivity

We Believe That Global Mining Productivity Will Continue to Decline Due to the Underlying Geological Deterioration; We Do Not Think the Potential Increase in Mining Technology Would Be Able to Reverse the Productivity Trend



In our base-case price forecast, we do not model any capital discipline and simply have new projects being commissioned as and when this becomes technically feasible. However, we note that the degree of technical feasibility for projects is often (very) difficult to assess. It can lead to some over-optimistic views on just how quickly new capacity can be brought on line. Furthermore, if we look at the level of depletion from existing sources of supply, it is severe. Consequently, by accounting for both the existing operating assets and those projects that already in production, we arrive at a picture as shown in Exhibit 208. We expect a 15Mt supply gap to open up by 2030 in mined copper. This imbalance is purely notional. When we look at the history, bar movements in terminal market inventory, the market is always in balance, and periods of under- or oversupply only apply on a forward-looking basis. Of course, price adjustments serve to close the gap between apparent supply and demand with the price level required to effect any change in these balances being determined by the underlying industry cost structure. However, for the next 18 months or so, it is clear that the market is in balance to a slight oversupply. Beyond this, a gap appears to open up and this gap ought to put upward pressure on price. The question is, of course, how this change in price will manifest itself in the supply and demand of copper.

Exhibit 208 In Aggregate, We See a Balanced to Slightly Oversupplied Copper Market for the Next 18 to 24 Months (Assuming No Further Supply Side Shocks or Demand-Side Weakness); However, Post This Period, Current Capacity Falls Well Short of Demand


We see four mechanisms at work here: Demand will be shed as high prices either deter the consumption of metal

altogether or encourage the consumption of lower-priced metals (i.e., substitution out of copper into aluminum).

Scrap and secondary consumption will rise as price encourages faster recycling rates and more effective reclamation of metal.

Short-term mine output will rise as changes to short-term mine plans are incentivized (i.e., mines exploiting low-grade material that would have closed at lower prices can continue producing for a few more years). In addition, high-grade mines are more likely to be front-loading exploitation of above-average grade zones of mineralization, given the often mistaken belief that a price fall will

0

5,000

10,000

15,000

20,000

25,000

30,000

Min

ed C

op

per

-k

t

Supply Demand Balance from Existing Operations and Projects in Construction

SxEw Concentrate Demand

The Difficulty in Assessing Technical Feasibility Is Something That Can Lead to Very Different Forecasts of Future Copper Supply

There Are Four Mechanisms That Explain the Interplay Between the Change in Price and the Underlying Supply and Demand of Copper, With Only the ''Approval of New Projects'' Mechanism Having the Capacity to Alter the Cost Curve Structure



subsequently follow and front-loading volumes therefore represents a value-maximizing strategy.

New projects will be approved. Only the last one of these mechanisms has the capacity to alter the structure of

the cost curve significantly and permanently. Consequently, it is the mechanism that is most often scrutinized in price analysis. However, new project execution is the most uncertain of all the means by which equilibrium is established, as it is the most geared to long-term future price expectations. It is clear that there is plenty of new potential copper capacity from a fundamental geological perspective (see Exhibit 209). The issue, however, is not that the world cannot supply more copper, but the price that will be required to incentivize this. The incentive curve for new copper projects suggests that many of them will fail to make an adequate return, if invested at today's prices. Moreover, there is further difficulty in getting boards to approve new investments at a time of falling or flat prices and in the face of increasingly vociferous calls for capex discipline and returns of capital to shareholders. When we factor in these variables, the picture seen in Exhibit 210 emerges. The cyclical nature of price and investment in new projects is clear. In order to close the 2020 gap, new investment will have to be made in copper and today's environment makes that difficult. As the supply/demand gap begins to make itself felt, prices will rise to ensure that the gap is never actually seen. We expect this rise in price to bring the new wave of fresh mine capacity on line. However, it is difficult to see how this dynamic can play out if (as consensus has it) prices continue to decline monotonically, yet productivity falls and costs rise.

When we look at the geographic origin of new supply, we see a number of interesting contrasts (see Exhibit 211). The rise of Africa as a key component in global supply. Increasingly, the world

will come to rely on Zambian and DRC copper supply. However, this also comes with a risk and the compensation for this risk must imply a slower rate of capitalization of African geology than would be the case if that risk were not present (see Exhibit 212).

The struggle to continue growth in Chile versus the rise of Peru. Given the level of current Chilean copper production, we deem it very unlikely that Chile will be able to grow overall production significantly, absent significant capital investment. At best, we expect it to be able to keep pace with depletion and maintain its current position. In contrast, we believe that Peru offers real growth potential assuming that the political environment does not deteriorate (see Exhibit 213). This highlights the move out of the "safe" mining jurisdictions into increasingly untested new frontier geographies.



Exhibit 209 There Is a Sufficient "Reservoir" of New Projects to Fill Any Supply/demand Gap; the Only Question Is the Price Required to Incentivize These Projects to Come on Line


Exhibit 210 The Current Low Price for Copper (and Copper By-Products Such as Gold and Molybdenum) Makes It Very Hard for New Projects to Get Approved, Thus Opening up a Supply Gap in the 2015E-20E Period


0

5000

10000

15000

20000

25000

30000

35000

40000

Existing Mined Supply by 2030 Possible New Projects Mined Copper Demand by 2030

Hypothetical Surplus by 2030

Min

ed O

op

per

-kt

Potential New Supply by 2030

0

5,000

10,000

15,000

20,000

25,000

30,000

Min

ed C

op

per

-kt

Supply Demand Balance With as Yet Unapproved Projects

SxEw Concentrate Unapproved SxEw Unapproved Conc Demand



Exhibit 211 We See the U.S. as One of the Biggest Winners (Along With Peru) in the Supply of Future Copper


Exhibit 212 However, African Growth Is an Absolute Requirement for Longer-Term Copper Supply


0

5,000

10,000

15,000

20,000

25,000

30,000

Min

ed

Co

pp

er -

ktMined Supply by Location

Chile USA China Peru Zambia DRC Canada Australia Mexico Mongolia Other

0

200

400

600

800

1,000

1,200

1,400

1,600

1,800

2005 2006 2007 2008 2009 2010 2011 2012 2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E

Min

ed C

op

pe

r -k

t

"New World" Copper — DRC and Zambia

Zambia DRC



Exhibit 213 We Are Getting Increasingly Worried About the Longer-Term Ability of Chile to Maintain Its Capacity


0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

2005 2006 2007 2008 2009 2010 2011 2012 2013E 2014E 2015E 2016E 2017E 2018E 2019E 2020E

Min

ed C

op

pe

r -k

t"Old World" Copper Chile and Peru

Chile Peru



Cost Curve Dynamics — The Evolution of the Cost Structure of Supply We summarize the net effect of our demand forecast, coupled to the incentive and cash cost structure of the industry, in Exhibit 214. It shows how the marginal unit of supply is expected to move over time. The dynamic represents an interplay between differential cost escalations in mining versus the net supply/demand balance.

Exhibit 214 We Expect Copper Price to Remain High for as Long as China's Wage Inflation

Grows Faster Than Its Mining Productivity, Unless the Western Miners Expend So Much Capital as to Displace the Need for Such Marginal Activity Entirely


-

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

20,000

0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% 55% 60% 65% 70% 75% 80% 85% 90% 95%100%

Cas

h C

ost

s C

1+S

IB C

apex

-U

S$/

t -N

om

inal

Cumulative Mined Production as % of Demand

2012 2017 2022



Our ''Base-Case'' Price Forecast Exhibit 215 shows our "base-case" price forecast. We use the evolving dynamics of how the marginal unit of supply will be sourced to anchor our forecast. We are more bullish than consensus over the whole of the forecast period as we believe that the real driver of the copper price "super-cycle" is the Chinese mining productivity. We believe that non-Chinese mine supply will struggle to maintain current output levels and, as such, we are optimistic about the prospects for this commodity even if an understanding of the need for greater capital discipline does not emerge in the major mining houses. Naturally, should we see greater capital discipline emerge from the mining majors, we would become incrementally more positive on the outlook for copper as well as other commodities.

Exhibit 215 We See a Supply-Demand Balance in the Immediate Short Term; We Expect the Market to Become Increasingly Stretched Over the Next Two Years; by Then We Expect Cost Escalation and Productivity in China (and Elsewhere) to Have Only Deteriorated Further, Thus Leading to Ever Higher Prices Necessary to Incentivize the Next Wave of Mine Investment


0

2,000

4,000

6,000

8,000

10,000

12,000

Jan

-04

Jun

-04

Nov-0

4

Apr-

05

Sep

-05

Fe

b-0

6

Jul-06

Dec-0

6

May-0

7

Oct-

07

Mar-

08

Aug

-08

Jan

-09

Jun

-09

Nov-0

9

Ap

r-1

0

Sep

-10

Fe

b-1

1

Jul-11

Dec-1

1

May-1

2

Oct-

12

Mar-

13

Aug

-13

Jan

-14

Jun

-14

Nov-1

4

Apr-

15

Sep

-15

Fe

b-1

6

Jul-16

Dec-1

6

May-1

7

Oct-

17

Mar-

18

Aug

-18

Jan

-19

Jun

-19

Nov-1

9

Ap

r-2

0

Sep

-20

Co

pp

er

Pri

ce

-U

S$

/t

SCB vs. Consensus and the Forward Curve

Cu Spot Cu Forward Cu Consensus Cu SCB

We Believe That the Real Driver of the Copper Price ''Super-Cycle'' Is the Chinese Mining Productivity, and We See China Struggling to Maintain Current Output Levels



Appendix

Appendix: Top 10 Mines Account for a Quarter of Global Copper Supply To powerfully end the supply side portion of the Blackbook, we wanted to shed some light on the 10 largest copper mines globally, of which eight are located in Chile and one each in Peru and Indonesia. The development of these mines has been instrumental in securing cheap copper supply over the last few decades.

So far, we have looked at the development of the world's copper supply on a country-by-country basis. However, there is significant granularity within any country's mine supply, with a finite number of assets contributing to overall output. Moreover, a very significant difference exists between the average copper asset and the handful of truly "Tier 1" operations that have stood behind the increases in mined production over the last few decades. The top 10 mines (or top 1.5% mines out of ~640 known copper operations) account for 25% of mined copper supply (see Exhibit 216). These mines are hugely influential in determining the future copper price — not because they set the supply of the marginal ton themselves, but because they determine the requirement for marginal units of supply. Consequently, a familiarity with these operations is essential for any view on the future of the copper price (see Exhibit 217). The appendix offers an overview of the 10 largest copper mines (see Exhibit 220 through Exhibit 289).

We also present the 10 largest new copper projects, which account for 18% of possible new supply (see Exhibit 218). The potential incremental recovered copper is categorized project by project in Exhibit 219. Given the concentration of new sources of supply in a handful of very large projects, we also provide further details on each one of them as a guide to the most important new supply side developments (see Exhibit 290 through Exhibit 359).

Exhibit 216 1.5% of Mines Account for 25% of Global

Copper Supply

Exhibit 217 Understanding the Development of These 10 Mines Is Critical in Understanding the Forward-Looking Copper Balance


25%

75%

Supply from Top 10 Mines Globally

Top 10 Bottom 630

0

200

400

600

800

1,000

1,200

kt C

u

10 Largest Copper Mines in 2012

Just 10 Mines Account for 25% of Global Supply; We See a Similar Pattern Replicated in New Project Development — 10 Projects Represent 18% of Potential New Supply



Exhibit 218 3% of New Projects Account for 18% of Possible New Supply

Exhibit 219 It Is Rio Tinto and Glencore Xstrata That Have the Greatest Exposure to These Developments

Source: MEG and Bernstein analysis. Source: MEG and Bernstein analysis.

18%

82%

Top 10 Copper Projects by Recovered Cu

Top 10 Bottom 317

0

100

200

300

400

500

600

700

800

10 Largest Projects by Recovered Cu



Escondida Exhibit 220 Escondida Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 221 Escondida Ownership Exhibit 222 Escondida Metal Exposure

Source: MEG and Bernstein estimates and analysis. Source: MEG and Bernstein estimates and analysis.

General Information Details/FactsCountry ChileLocation 170km SE of AntofagastaState/Province AntofagastaLocale Atacama Desert In N. ChileStart Up 1990 (Q4)Commodity Copper/Gold/SilverDevelopment Stage ProductionMine Type Open PitLatitude/Longitude 24°16'0" S, 69°4'0" W

Geology Details/FactsZone Name EscondidaOre Genesis Supergene (Secondary) Enrichment Hydrothermal processesOrebody Type Porphyry DepositOre Mineral Chalcocite, Covellite, Chalcopyrite, Pyrite, BorniteClass of Ore Oxide, SulfideOre Controls FaultingWidth 2.5 kmLength 4.5 kmThickness 600 mHost Rock Andesite (Paleocene)Country Rock Sedimentary (Mesozoic), Volcanics (Paleozoic)Strike N/A

57.5%30.0%

8.3%

3.0% 1.3%

Ownership of Escondida

BHP Billiton Rio Tinto Mitsubishi Corp

Nippon Mining Mitsubishi Minerals

97%

1% 2%

Escondida Metal-by-Metal Revenue

Cu Ag Au



Exhibit 223 Escondida Geological Endowment Exhibit 224 Escondida Ore Grade

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 225 Escondida Head Grade Exhibit 226 Escondida Cost Position


0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

Reserves Resources

Escondida Reserves & Resources (Mt)

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

Reserves Resources

Escondida Reserves & Resources Cu Grade (%)

0.60

0.80

1.00

1.20

1.40

1.60

1.80

2.00

200

020

01

200

220

03

200

420

05

200

620

07

200

820

09

201

020

11

201

22

013E

201

4E2

015E

201

6E2

017E

201

8E2

019E

202

0E

Escondida Cu Mill Head Grade (%)

10,000

9,000

12,000

11,000

8,000

7,000

6,000

5,000

4,000

3,000

2,000

1,000

0

20,00015,00010,0000 5,000



Escondida



Antamina Exhibit 227 Antamina Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 228 Antamina Ownership Exhibit 229 Antamina Metal Exposure


General Information Details/FactsCountry PeruLocation 280 km N of Lima; 135 km NE of HuarmeyState/Province Ancash (District/Town is San Marcos)Locale Andes MountainsStart Up 2001 (Q4)Commodity Copper/Zinc/Molybdenum/Lead/Silver/BismuthDevelopment Stage ProductionMine Type Open PitLatitude/Longitude 9°32'21" S, 77°3'0" W

Geology Details/FactsZone Name AntaminaOre Genesis N/AOrebody Type N/AOre Mineral Chalcopyrite, Sphalerite, Bornite, Pyrite, MagnetiteClass of Ore N/AOre Controls N/AWidth 1 kmLength 2.5 kmThickness N/AHost Rock Skarn (Tacite)Country Rock N/AStrike SW-NE

33.8%

33.8%

22.5%

10.0%

Ownership of Antamina

BHP Biliton Glencore Xstrata

Teck Resources Mitsubishi Corp

76%

0%

10%

4%

10%

Antamina Metal-by-Metal Revenue

Cu Pb Zn Mo Ag



Exhibit 230 Antamina Geological Endowment Exhibit 231 Antamina Ore Grade

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 232 Antamina Head Grade Exhibit 233 Antamina Cost Position


0

200

400

600

800

1,000

1,200

Reserves Resources

Antamina Reserves & Resources (Mt)

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

Reserves Resources

Antamina Reserves & Resources Cu Grade (%)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

200

02

001

200

22

003

200

42

005

200

62

007

200

82

009

201

02

011

201

220

13E

2014

E20

15E

2016

E20

17E

2018

E20

19E

2020

E

Antamina Cu Mill Head Grade (%)

8,000

7,000

6,000

5,000

4,000

3,000

2,000

1,000

0

20,00015,00010,0005,0000

10,000

9,000

11,000

12,000C1 costs + SIB Capex –US$/t


Antamina



Los Pelambres Exhibit 234 Los Pelambres Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 235 Los Pelambres Ownership Exhibit 236 Los Pelambres Metal Exposure


General Information Details/FactsCountry ChileLocation 46 km E of Salamanca; 200 km N of SantiagoState/Province CoquimboLocale Andes MountainsStart Up 1999 (Q4)Commodity Copper/Molybdenum/Gold/SilverDevelopment Stage ProductionMine Type Open PitLatitude/Longitude 31°43'4" S, 70°29'22" W

Geology Details/FactsZone Name Los PelambresOre Genesis N/AOrebody Type Porphyry DepositOre Mineral Chalcocite, Chalcopyrite, BorniteClass of Ore N/AOre Controls N/AWidth N/ALength N/AThickness N/AHost Rock Andesite, DioriteCountry Rock N/AStrike N/A

60.0%16.3%

10.0%

8.8%5.0%

Ownership of Los Pelambres

Antofagasta Plc Pan Pacific Copper Co Ltd

Mitsubishi Materials Corp Marubeni Corp

Mitsubishi Corp

87%

8%2% 3%

Los Pelambres Metal-by-Metal Revenue

Cu Mo Ag Au



Exhibit 237 Los Pelambres Geological Endowment Exhibit 238 Lost Pelambres Ore Grade

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 239 Los Pelambres Head Grade Exhibit 240 Los Pelambres Cost Position


0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

5,000

Reserves Resources

Los Pelambres Reserves & Resources (Mt)

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

Reserves Resources

Los Pelambres Reserves & Resources Cu Grade (%)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

200

02

001

200

22

003

200

42

005

200

62

007

200

82

009

201

02

011

201

220

13E

2014

E20

15E

2016

E20

17E

2018

E20

19E

2020

E

Los Pelambres Cu Mill Head Grade (%)12,000

11,000

10,000

9,000

8,000

7,000

6,000

5,000

4,000

3,000

2,000

1,000

0

20,00015,00010,0005,0000



Los Pelambres



El Teniente Exhibit 241 El Teniente Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 242 El Teniente Ownership Exhibit 243 El Teniente Metal Exposure


General Information Details/FactsCountry ChileLocation 44 km NE of Rancagua, 80 Km SE of SantiagoState/Province Region lV (District/Town O'Higgins [Region VI])/RancaguaLocale N/AStart Up 1904Commodity Copper/Molybdeum/Gold/SilverDevelopment Stage ProductionMine Type UndergroundLatitude/Longitude 34°4'59" S, 70°22'0" W

Geology Details/FactsZone Name N/AOre Genesis N/AOrebody Type Porphyry DepositOre Mineral N/AClass of Ore N/AOre Controls BrecciationWidth N/ALength N/AThickness N/AHost Rock Intrusive (plutonic)Country Rock N/AStrike N/A

100%

Ownership of El Teniente

Codelco

93%

4% 2% 1%

El Teniente Metal-by-Metal Revenue

Cu Mo Ag Au



Exhibit 244 El Teniente Geological Endowment Exhibit 245 El Teniente Ore Grade

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 246 El Teniente Head Grade Exhibit 247 El Teniente Cost Position


0

500

1,000

1,500

2,000

2,500

3,000

Reserves Resources

El Teniente Reserves & Resources (Mt)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

Reserves Resources

El Teniente Reserves & Resources Cu Grade (%)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

E20

14E

2015

E20

16E

2017

E20

18E

2019

E20

20E

El Teniente Cu Mill Head Grade (%)

0

12,000

11,000

10,000

9,000

8,000

7,000

6,000

5,000

4,000

3,000

2,000

1,000

0

20,00015,00010,0005,000



El Teniente



Chuquicamata Exhibit 248 Chuquicamata Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 249 Chuquicamata Ownership Exhibit 250 Chuquicamata Metal Exposure


General Information Details/FactsCountry ChileLocation 16 km N of Calama; 1,592 km N of SantiagoState/Province AntofagastaLocale Atacama Desert; northern ChileStart Up 1910Commodity Copper/Molybdenum/Gold/Silver/RheniumDevelopment Stage ProductionMine Type Open Pit/Tailings/UndergroundLatitude/Longitude 22°17'30" S, 68°54'30" W

Geology Details/FactsZone Name N/AOre Genesis N/AOrebody Type N/AOre Mineral N/AClass of Ore N/AOre Controls N/AWidth N/ALength N/AThickness N/AHost Rock N/ACountry Rock N/AStrike N/A

100%

Ownership of Chuquicamata

Codelco

90%

5%4% 1%

Chuquicamata Metal-by-Metal Revenue

Cu Mo Ag Au



Exhibit 251 Chuquicamata Geological Endowment Exhibit 252 Chuquicamata Ore Grade

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 253 Chuquicamata Head Grade Exhibit 254 Chuquicamata Cost Position


0

200

400

600

800

1,000

1,200

1,400

1,600

1,800

Reserves Resources

Chuquicamata Reserves & Resources (Mt)

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

Reserves Resources

Chuquicamata Reserves & Resources Cu Grade (%)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

200

02

001

200

22

003

200

42

005

200

62

007

200

82

009

201

02

011

201

220

13E

2014

E20

15E

2016

E20

17E

2018

E20

19E

2020

E

Chuquicamata Cu Mill Head Grade (%)

0

20,00015,00010,0005,0000

8,000

7,000

6,000

5,000

4,000

3,000

2,000

1,000

12,000

11,000

10,000

9,000



Chuqui



Grasberg Exhibit 255 Grasberg Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 256 Grasberg Ownership Exhibit 257 Grasberg Metal Exposure


General Information Details/FactsCountry IndonesiaLocation N/AState/Province Papua (Town/District Timika/Jaya Wijaya Mountains)Locale N/AStart Up 1972Commodity Copper/Gold/SilverDevelopment Stage ProductionMine Type Open Pit and UndergroundLatitude/Longitude 4°7'59" S, 137°40'0" E

Geology Details/FactsZone Name GrasbergOre Genesis N/AOrebody Type Porphyry Deposit, SkarnOre Mineral N/AClass of Ore N/AOre Controls N/AWidth N/ALength N/AThickness N/AHost Rock Limestone, Monzonite, GranodioriteCountry Rock N/AStrike N/A

90.6%

9.4%

Ownership of Grasberg

Freeport-Mcmoran Copper and Gold Inc

Government of Indonesia

61%

2%

37%

Grasberg Metal-by-Metal Revenue

Cu Ag Au



Exhibit 258 Grasberg Geological Endowment Exhibit 259 Grasberg Ore Grade

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 260 Grasberg Head Grade Exhibit 261 Grasberg Cost Position


0

500

1,000

1,500

2,000

2,500

3,000

Reserves Resources

Grasberg Reserves & Resources (Mt)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

Reserves Resources

Grasberg Reserves & Resources Cu Grade (%)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

200

02

001

200

22

003

200

42

005

200

62

007

200

82

009

201

02

011

201

220

13E

2014

E20

15E

2016

E20

17E

2018

E20

19E

2020

E

Grasberg Cu Mill Head Grade (%)

5,000

4,000

3,000

2,000

1,000

0

20,00015,00010,0005,0000

12,000

11,000

10,000

9,000

8,000

7,000

6,000



Grasberg



Los Bronces Exhibit 262 Los Bronces Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 263 Los Bronces Ownership Exhibit 264 Los Bronces Metal Exposure


General Information Details/FactsCountry ChileLocation 45 km NE of SantiagoState/Province Valparaiso (Town Santiago)Locale N/AStart Up 1925Commodity Copper/MolybdenumDevelopment Stage ProductionMine Type Open PitLatitude/Longitude 33°8'56" S, 70°16'54" W

Geology Details/FactsZone Name Los BroncesOre Genesis Hydrothermal processesOrebody Type Breccia Fill, StockworkOre Mineral Chalcopyrite, Specularite, MolybdeniteClass of Ore N/AOre Controls N/AWidth N/ALength N/AThickness N/AHost Rock Intrusive (plutonic), VolcanicsCountry Rock N/AStrike N/A

50.1%

24.5%

20.4%

5.0%

Ownership of Los Bronces

Anglo American Codelco

Mitsubishi Corp Anglo American Sur

94%

1%2%

3%

Los Bronces Metal-by-Metal Revenue

Cu Mo Ag Au



Exhibit 265 Los Bronces Geological Endowment Exhibit 266 Los Bronces Ore Grade

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 267 Los Bronces Head Grade Exhibit 268 Los Bronces Cost Position


0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

5,000

Reserves Resources

Los Bronces Reserves & Resources (Mt)

0.00

0.10

0.20

0.30

0.40

0.50

0.60

Reserves Resources

Los Bronces Reserves & Resources Cu Grade (%)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

201

3E2

014E

201

5E2

016E

201

7E2

018E

201

9E2

020E

Los Bronces Cu Mill Head Grade (%)

7,000

6,000

5,000

4,000

3,000

2,000

1,000

0

20,00015,00010,0005,0000

12,000

11,000

10,000

9,000

8,000



Los Bronces



Radomiro Tomic SxEw Exhibit 269 Radomiro Tomic Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 270 Radomiro Tomic Ownership Exhibit 271 Radomiro Tomic Metal Exposure


General Information Details/FactsCountry ChileLocation 35 km N of Calama; 6 km N of the Chuquicamata mineState/Province AtacamaLocale Atacama Desert; northern ChileStart Up 1998 (Q1)Commodity Copper/MolybdenumDevelopment Stage ProductionMine Type Open PitLatitude/Longitude 22°13'59" S, 68°55'0" W

Geology Details/FactsZone Name Radomiro TomicOre Genesis N/AOrebody Type N/AOre Mineral AtacamiteClass of Ore Oxide, SulfideOre Controls N/AWidth N/ALength N/AThickness N/AHost Rock N/ACountry Rock N/AStrike N/A

100%

Ownership of Radomiro Tomic

Codelco

100%

Radomiro Tomic Metal-by-Metal Revenue

Cu



Exhibit 272 Radomiro Tomic Geological Endowment Exhibit 273 Radomiro Tomic Ore Grade

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 274 Radomiro Tomic Leach Grade Exhibit 275 Radomiro Tomic Cost Position


0

100

200

300

400

500

600

700

800

900

Reserves Resources

Radomiro Tomic Reserves & Resources (Mt)

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

Reserves Resources

Radomiro Tomic SX-EW Reserves & Resources Cu Grade (%)

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

200

02

001

200

22

003

200

42

005

200

62

007

200

82

009

201

02

011

201

220

13E

2014

E20

15E

2016

E20

17E

2018

E20

19E

2020

E

Radomiro SX-EW Leach Head Grade (%)

20,00015,00010,0005,0000

12,000

11,000

10,000

9,000

8,000

7,000

6,000

5,000

4,000

3,000

2,000

1,000

0



Radomiro Tomic



Andina Exhibit 276 Andina Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 277 Andina Ownership Exhibit 278 Andina Metal Exposure


General Information Details/FactsCountry ChileLocation 40 km NE of SantiagoState/Province Valparaiso (Town Santiago)Locale Mt Aconcagua; Fifth RegionStart Up 1970Commodity Copper/Molybdenum/Gold/SilverDevelopment Stage ProductionMine Type Underground and Open PitLatitude/Longitude 33°9'5" S, 70°15'21" W

Geology Details/FactsZone Name Andina DivisionOre Genesis N/AOrebody Type Porphyry DepositOre Mineral Chalcopyrite, MolybdeniteClass of Ore N/AOre Controls BrecciationWidth N/ALength N/AThickness N/AHost Rock Intrusive (plutonic), VolcanicsCountry Rock N/AStrike N/A

100%

Ownership of Andina

Codelco

90%

6% 3% 1%

Andina Metal-by-Metal Revenue

Cu Mo Ag Au



Exhibit 279 Andina Geological Endowment Exhibit 280 Andina Ore Grade

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 281 Andina Head Grade Exhibit 282 Andina Cost Position


0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

Reserves Resources

Andina Reserves & Resources (Mt)

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

Reserves Resources

Andina Reserves & Resources Cu Grade (%)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

201

3E2

014E

201

5E2

016E

201

7E2

018E

201

9E2

020E

Andina Mill Cu Head Grade (%)

9,000

8,000

7,000

6,000

5,000

4,000

3,000

2,000

1,000

0

20,00015,00010,0005,0000

12,000

11,000

10,000



Andina



Collahuasi Exhibit 283 Collahuasi Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 284 Collahuasi Ownership Exhibit 285 Collahuasi Metal Exposure


Country ChileLocation 160 km SE of the port of IquiqueState/Province Tarapaca (Town Iquique)Locale Andes Mountains; Northern ChileStart Up 1999 (Q1)Commodity Copper/Molybdenum/SilverDevelopment Stage ProductionMine Type Open PitLatitude/Longitude 20°59'21" S, 68°38'9" W

Geology Details/FactsZone Name CollahuasiOre Genesis Hydrothermal processes; ReplacementOrebody Type Porphyry DepositOre Mineral Chalcocite, Chalcopyrite, Bornite, CovelliteClass of Ore Sulfide, OxideOre Controls Fracturing, Vein (Lode)Width N/ALength N/AThickness N/AHost Rock Intrusive (plutonic)Country Rock N/AStrike N/A

44%

44%

8%

4%

Ownership of Collahuasi

Glencore Anglo Mitsui Nippon

90%

4%4% 2%

Collahuasi Metal-by-Metal Revenue

Cu Mo Ag Au



Exhibit 286 Collahuasi Geological Endowment Exhibit 287 Collahuasi Ore Grade

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 288 Collahuasi Head Grade Exhibit 289 Collahuasi Cost Position


0

500

1,000

1,500

2,000

2,500

3,000

3,500

Reserves Resources

Collahuasi Reserves & Resources (Mt)

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

Reserves Resources

Collahuasi Reserves & Resources Cu Grade (%)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

2.00

200

020

01

200

220

03

200

420

05

200

620

07

200

820

09

201

020

11

201

22

013E

201

4E2

015E

201

6E2

017E

201

8E2

019E

202

0E

Collahuasi Cu Mill Head Grade (%)

1,000

0

20,00015,00010,0005,0000

12,000

11,000

10,000

9,000

8,000

7,000

6,000

5,000

4,000

3,000

2,000



Collahuasi



Oyu Tolgoi Exhibit 290 Oyu Tolgoi Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 291 Oyu Tolgoi Ownership Exhibit 292 Oyu Tolgoi Metal Exposure


General Information Details/FactsCountry MongoliaLocation 550 km S of Ulaanbaatar; 80 km N of the Chinese borderState/Province OmnogoviLocale South Gobi region of Southern MongoliaStart Up 2013 (Q1)Commodity Copper/Gold/Silver/MolybdenumDevelopment Stage ProductionMine Type Open Pit, Underground, Stock PileLatitude/Longitude 43°1'0" N, 106°51'0" E

Geology Details/FactsZone Name Tuquoise HillOre Genesis N/AOrebody Type Porphyry DepositOre Mineral Chalcopyrite, Bornite, Magnetite, ChalcociteClass of Ore SulfideOre Controls BrecciationWidth N/ALength N/AThickness N/AHost Rock Andesite (Devonian), MonzodioriteCountry Rock N/AStrike N/A

34%

34%

32%

Ownership of Oyu Tolgoi

Rio Tinto Government of Mongolia Other

81%

2%

17%

Oyu Tolgoi Expected Revenue (Y2025) Metal-by-Metal Split

Cu Ag Au



Exhibit 293 Oyu Tolgoi Grade Profile Exhibit 294 Oyu Tolgoi Production Profile

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 295 Oyu Tolgoi Geological Endowment Exhibit 296 Oyu Tolgoi Grade


0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

Oyu Tolgoi Cu Mill Head Grade (%)

0

100

200

300

400

500

600

700

800

900

201

0

201

2

201

4E

201

6E

201

8E

202

0E

202

2E

202

4E

202

6E

202

8E

203

0E

203

2E

203

4E

203

6E

203

8E

204

0E

kt C

u

Oyu Tolgoi Production Profile

0

500

1,000

1,500

2,000

2,500

Reserves Resources

Oyu Tolgoi Reserves & Resources (Mt)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

Reserves Resources

Oyu Tolgoi Reserves & Resources Cu Grade (%)



Resolution Exhibit 297 Resolution Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 298 Resolution Ownership Exhibit 299 Resolution Metal Exposure


General Information Details/FactsCountry United StatesLocation 145 km E of PhoenixState/Province ArizonaLocale N/AStart Up 2023Commodity Copper/Gold/MolybdenumDevelopment Stage FeasibilityMine Type Underground - Block & Panel CavingLatitude/Longitude 33°18'18" N, 111°3'52" W

Geology Details/FactsZone Name Superior MineOre Genesis Hydrothermal processesOrebody Type Porphyry DepositOre Mineral Chalcopyrite, Chalcocite, Bornite, Gold, Enargite, TennantiteClass of Ore Sulfide, NativeOre Controls Stratigraphy, FoldingWidth 590 m (average)Length 1200 m (average)Thickness 6 m (average)Host Rock Limestone (Precambrian), Monzonite, Diabase (Precambrian)Country Rock N/AStrike North 00 degrees South

55%

45%

Ownership of Resolution Copper

Rio Tinto BHP Billiton

90%

3%

3% 4%

Resolution Copper Expected Revenue (Y2025) Metal-by-Metal Split

Cu Mo Ag Au



Exhibit 300 Resolution Grade Profile Exhibit 301 Resolution Production Profile

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 302 Resolution Geological Endowment Exhibit 303 Resolution Grade


0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

Resolution Copper Cu Mill Head Grade (%)

0

100

200

300

400

500

600

201

0

201

2

2014

E

2016

E

2018

E

2020

E

2022

E

2024

E

2026

E

2028

E

2030

E

2032

E

2034

E

2036

E

2038

E

2040

E

kt C

u

Resolution Production Profile

0

200

400

600

800

1,000

1,200

1,400

1,600

1,800

Reserves Resources

Resolution Copper Reserves & Resources (Mt)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

Reserves Resources

Resolution Copper Reserves & Resources Cu Grade (%)



La Granja Exhibit 304 La Granja Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 305 La Granja Ownership Exhibit 306 La Granja Metal Exposure


General Information Details/FactsCountry PeruLocation 650 km N of LimaState/Province ChotaLocale Northern PeruStart Up 2017 (Q4)Commodity Copper/Gold/Silver/ZincDevelopment Stage Reserves DevelopmentMine Type Open PitLatitude/Longitude 6°21'23" S, 79°7'0" W

Geology Details/FactsZone Name La GranjaOre Genesis N/AOrebody Type Disseminated, Porphyry Deposit, SkarnOre Mineral Chalcopyrite, Bornite, Chalcocite, CovelliteClass of Ore N/AOre Controls FaultingWidth N/ALength N/AThickness N/AHost Rock Skarn (Tacite)Country Rock Limestone, Siltstone, QuartziteStrike N/A

100%

Ownership of La Granja SX-EW

Rio Tinto

100%

La Granja Expected Revenue (Y2025) Metal-by-Metal Split

Cu



Exhibit 307 La Granja Grade Profile Exhibit 308 La Granja Production Profile

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 309 La Granja Geological Endowment Exhibit 310 La Granja Grade


0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

La Granja SX-EW Cu Leach Head Grade (%)

0

50

100

150

200

250

300

350

201

0

201

2

2014

E

2016

E

2018

E

2020

E

2022

E

2024

E

2026

E

2028

E

2030

E

2032

E

2034

E

2036

E

2038

E

2040

E

kt C

u

La Granja Production Profile

0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

Reserves Resources

La Granja Reserves & Resources (Mt)

0.00

0.10

0.20

0.30

0.40

0.50

0.60

Reserves Resources

La Granja Reserves & Resources Cu Grade (%)



Tampakan Exhibit 311 Tampakan Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 312 Tampakan Ownership Exhibit 313 Tampakan Metal Exposure


General Information Details/FactsCountry PhilippinesLocation about 14 km E of Tampakan; 55 km NNW of General SantosState/Province South CotabatoLocale N/AStart Up 2019Commodity Copper/Gold/Molybdenum/ArsenicDevelopment Stage FeasibilityMine Type Open PitLatitude/Longitude 6°28'32" N, 125°3'22" E

Geology Details/FactsZone Name TampakanOre Genesis EpithermalOrebody Type Porphyry DepositOre Mineral Chalcopyrite, Bornite, Pyrite, MolybdeniteClass of Ore SulfideOre Controls N/AWidth N/ALength N/AThickness N/AHost Rock N/ACountry Rock N/AStrike Andesite

63%

38%

Ownership of Tampakan

Glencore Xstrata Indophil Resources NL

81%

19%

Tampakan Expected Revenue (Y2025) Metal-by-Metal Split

Cu Au



Exhibit 314 Tampakan Grade Profile Exhibit 315 Tampakan Production Profile

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 316 Tampakan Geological Endowment Exhibit 317 Tampakan Grade


0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

Tampakan Cu Mill Head Grade (%)

0

50

100

150

200

250

300

350

400

450

500

201

0

201

2

2014

E

2016

E

2018

E

2020

E

2022

E

2024

E

2026

E

2028

E

2030

E

2032

E

2034

E

2036

E

2038

E

2040

E

kt C

u

Tampakan Production Profile

0

500

1,000

1,500

2,000

2,500

3,000

Reserves Resources

Tampakan Reserves & Resources (Mt)

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

Reserves Resources

Tampakan Reserves & Resources Cu Grade (%)



Escondida Ph VI Exhibit 318 Escondida Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 319 Escondida Ownership Exhibit 320 Escondida Metal Exposure


General Information Details/FactsCountry ChileLocation 170km SE of AntofagastaState/Province AntofagastaLocale Atacama Desert in Norther ChileStart Up 1990 (Q4)Commodity Copper/Gold/SilverDevelopment Stage ProductionMine Type Open PitLatitude/Longitude 24°16'0" S, 69°4'0" W

Geology Details/FactsZone Name EscondidaOre Genesis Supergene (Secondary) Enrichment, Hydrothermal processesOrebody Type Porphyry DepositOre Mineral Chalcocite, Covellite, Chalcopyrite, Pyrite, BorniteClass of Ore Oxide, SulfideOre Controls FaultingWidth 2.5 kmLength 4.5 kmThickness 600 mHost Rock Andesite (Paleocene)Country Rock Sedimentary (Mesozoic), Volcanics (Paleozoic)Strike N/A

58%30%

8%

3% 1%

Ownership of Escondida

BHP Billiton Rio Tinto Mitsubishi Corp

Nippon Mining Mitsubishi Minerals

97%

1% 2%

Escondida (overall) Expected Revenue (Y2025) Metal-by-Metal Split

Cu Ag Au



Exhibit 321 Escondida Grade Profile Exhibit 322 Escondida Production Profile

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 323 Escondida Geological Endowment Exhibit 324 Escondida Grade


0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

Escondida (overall) Cu Mill Head Grade (%)

0

200

400

600

800

1,000

1,200

2000

2003

2006

2009

2012

201

5E

201

8E

202

1E

202

4E

202

7E

203

0E

203

3E

203

6E

203

9E

kt C

u

Escondida Production Profile

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

Reserves Resources

Escondida Reserves & Resources (Mt)

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

Reserves Resources

Escondida Reserves & Resources Cu Grade (%)



Golpu Exhibit 325 Golpu Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 326 Golpu Ownership Exhibit 327 Golpu Metal Exposure


General Information Details/FactsCountry Papua New GuineaLocation 65 km SW of Lae; 300 km N of Port MoresbyState/Province MorobeLocale N/AStart Up 2019 (Q1)Commodity Gold/Copper/Molybdenum/SilverDevelopment Stage Reserves DevelopmentMine Type UndergroundLatitude/Longitude 6°43'31" S, 146°37'45" E

Geology Details/FactsZone Name WafiOre Genesis Hydrothermal processesOrebody Type Porphyry Deposit, Vein (Lode)Ore Mineral N/AClass of Ore N/AOre Controls N/AWidth N/ALength N/AThickness N/AHost Rock Volcanics, SedimentaryCountry Rock N/AStrike N/A

50%50%

Ownership of Golpu

Newcrest Mining Harmony Gold Mining Co

78%

22%

Golpu Expected Revenue (Y2025) Metal-by-Metal Split

Cu Au



Exhibit 328 Golpu Grade Profile Exhibit 329 Golpu Production Profile

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 330 Golpu Geological Endowment Exhibit 331 Golpu Grade


0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

Golpu Head Mill Grade (%)

0

50

100

150

200

250

300

350

2010

2012

2014

E

2016

E

2018

E

2020

E

2022

E

2024

E

2026

E

2028

E

2030

E

2032

E

2034

E

2036

E

2038

E

2040

E

kt C

u

Golpu Production Profile

0

100

200

300

400

500

600

Reserves Resources

Golpu Reserves & Resources (Mt)

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

Reserves Resources

Golpu Reserves & Resources Cu Grade (%)



Las Bambas Exhibit 332 Las Bambas Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 333 Las Bambas Ownership Exhibit 334 Las Bambas Metal Exposure


General Information Details/FactsCountry PeruLocation 260 km SW of CuscoState/Province GrauLocale Inca Central AndeanStart Up 2015 (Q1)Commodity Copper/Molybdenum/Gold/SilverDevelopment Stage PreproductionMine Type Open Pit, UndergroundLatitude/Longitude 14°5'6" S, 72°28'2" W

Geology Details/FactsZone Name Las BambasOre Genesis N/AOrebody Type Skarn, Porphyry DepositOre Mineral N/AClass of Ore N/AOre Controls N/AWidth N/ALength N/AThickness N/AHost Rock N/ACountry Rock N/AStrike N/A

100%

Ownership of Las Bambas

Glencore Xstrata

66%5%

27%

2%

Las Bambas Expected Revenue (Y2025) Metal-by-Metal Split

Cu Mo Ag Au



Exhibit 335 Las Bambas Grade Profile Exhibit 336 Las Bambas Production Profile

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 337 Las Bambas Geological Endowment Exhibit 338 Las Bambas Grade


0.00

0.20

0.40

0.60

0.80

1.00

1.20

Las Bambas Cu Mill Head Grade (%)

0

100

200

300

400

500

600

2010

2012

2014

E

2016

E

2018

E

2020

E

2022

E

2024

E

2026

E

2028

E

2030

E

2032

E

2034

E

2036

E

2038

E

2040

E

kt C

u

Las Bambas Production Profile

0

200

400

600

800

1,000

1,200

1,400

1,600

1,800

Reserves Resources

Las Bambas Reserves & Resources (Mt)

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

Reserves Resources

Las Bambas Reserves & Resources Cu Grade (%)



Los Pelambres Expansion Exhibit 339 Los Pelambres Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 340 Los Pelambres Ownership Exhibit 341 Los Pelambres Metal Exposure


General Information Details/FactsCountry ChileLocation 46 km E of Salamanca; 200 km N of SantiagoState/Province CoquimboLocale Andes MountainsStart Up 1999 (Q4)Commodity Copper/Molybdenum/Gold/SilverDevelopment Stage ProductionMine Type Open PitLatitude/Longitude 31°43'4" S, 70°29'22" W

Geology Details/FactsZone Name Los PelambresOre Genesis N/AOrebody Type Porphyry DepositOre Mineral Chalcocite, Chalcopyrite, BorniteClass of Ore N/AOre Controls N/AWidth N/ALength N/AThickness N/AHost Rock Andesite, DioriteCountry Rock N/AStrike N/A

60%16%

10%

9%5%

Ownership of Los Pelambres

Antofagasta Plc Pan Pacific Copper Co Ltd

Mitsubishi Materials Corp Marubeni Corp

Mitsubishi Corp

90%

6%2%

2%

Los Pelambres (overall) Expected Revenue (Y2025) Metal-by-Metal

Cu Mo Ag Au



Exhibit 342 Los Pelambres Grade Profile Exhibit 343 Los Pelambres Production Profile

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 344 Los Pelambres Geological Endowment Exhibit 345 Los Pelambres Grade


0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

Los Pelambres (overall) Cu Mill Head Grade (%)

0

50

100

150

200

250

300

350

400

450

2000

2003

2006

2009

2012

201

5E

201

8E

202

1E

202

4E

202

7E

203

0E

203

3E

203

6E

203

9E

kt C

u

Los Pelambres Production Profile

0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

5,000

Reserves Resources

Los Pelambres Reserves & Resources (Mt)

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

Reserves Resources

Los Pelambres Reserves & Resources Cu Grade (%)



Pebble Exhibit 346 Pebble Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 347 Pebble Ownership Exhibit 348 Pebble Metal Exposure


General Information Details/FactsCountry United StatesLocation 330 km WSW of Anchorage; 27 km WSW of NondaltonState/Province AlaskaLocale SW AlaskaStart Up 2016Commodity Copper/Gold/Molybdenum/Silver/Palladium/RheniumDevelopment Stage Reserves DevelopmentMine Type Open Pit, UndergroundLatitude/Longitude 59°53'53" N, 155°17'44" W

Geology Details/FactsZone Name Pebble CopperOre Genesis Supergene (Secondary) Enrichment, Hydrothermal processesOrebody Type Porphyry DepositOre Mineral Chalcocite, Chalcopyrite, Bornite, MolybdenumClass of Ore Sulfide, OxideOre Controls N/AWidth N/ALength N/AThickness N/AHost Rock Argillite, SiltstoneCountry Rock Granodiorite

50%50%

Ownership of Pebble

Anglo American Northern Dynasty Minerals

61%

9%

2%

28%

Pebble Expected Revenue (Y2025) Metal-by-Metal Split

Cu Mo Ag Au



Exhibit 349 Pebble Grade Profile Exhibit 350 Pebble Production Profile

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 351 Pebble Geological Endowment Exhibit 352 Pebble Grade


0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

Pebble Cu Mill Head Grade (%)

0

50

100

150

200

250

300

350

400

2010

2012

2014

E

2016

E

2018

E

2020

E

2022

E

2024

E

2026

E

2028

E

2030

E

2032

E

2034

E

2036

E

2038

E

2040

E

kt C

u

Pebble Production Profile

0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

Reserves Resources

Pebble Reserves & Resources (Mt)

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

Reserves Resources

Pebble Reserves & Resources Cu Grade (%)



Andina Expansion Exhibit 353 Andina Overview

Source: MEG and Bernstein estimates and analysis. Exhibit 354 Andina Ownership Exhibit 355 Andina Metal Exposure


General Information Details/FactsCountry ChileLocation 40 km NE of SantiagoState/Province ValparaisoLocale Mt Aconcagua, Fifth RegionStart Up 1970Commodity Copper/Molybdenum/Gold/SilverDevelopment Stage ProductionMine Type Open Pit, UndergroundLatitude/Longitude 33°9'5" S, 70°15'21" W

Geology Details/FactsZone Name Andina DivisionOre Genesis N/AOrebody Type Porphyry DepositOre Mineral Chalcopyrite, MolybdeniteClass of Ore N/AOre Controls BrecciationWidth N/ALength N/AThickness N/AHost Rock Intrusive (Plutonic), VolcanicsCountry Rock N/AStrike N/A

100%

Ownership of Andina

Codelco

91%

5%

3% 1%

Andina (overall) Expected Revenue (Y2025) Metal-by-Metal Split

Cu Mo Ag Au



Exhibit 356 Andina Grade Profile Exhibit 357 Andina Production Profile

Source: Wood Mackenzie and Bernstein estimates and analysis. Source: Wood Mackenzie and Bernstein estimates and analysis. Exhibit 358 Andina Geological Endowment Exhibit 359 Andina Grade


0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

0.90

Andida (overall) Cu Mill Head Grade (%)

0

50

100

150

200

250

300

2000

2003

2006

2009

2012

201

5E

201

8E

202

1E

202

4E

202

7E

203

0E

203

3E

203

6E

203

9E

kt C

u

Andina Production Profile

0

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

Reserves Resources

Andina (overall) Reserves & Resources (Mt)

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0.70

0.80

Reserves Resources

Andina (overall) Reserves & Resources Cu Grade (%)



Index of Exhibits

1 Financial Overview 4

2 Copper Stands Out as Being the Most Overutilized Commodity Relative to Underlying Geological Endowment — Testimony to the Industrial Importance of This Metal and the Difficulty in Growing Supply in Anything Other Than a Supportive Price Environment 6

3 The Long-Term History of Copper Porphyry Discovery Had a Marked Impact on the Copper Price: Falling Prices Have Been Associated With Increased Finds of Relatively Few Massive Ore Bodies — We See No New Chile on the Horizon 7

4 Chile Stands Out as the Most Important Source of Supply, But This Was Not Always the Case — While Chile's Growth Has Slowed, China's Has Accelerated 7

5 A Corrected China Cost Curve Would Display Much More High-Cost Production: The Relationship Between Price and Chinese Copper Output Tells Us That They Are Not Unrelated Phenomena 8

6 It Is the Differential Cost Escalation That Drives Real Commodity Price Increases in USD Terms 9

7 "Grade Is King" and Global Head Grades Are Declining 10

8 Nearly 90% of the Movement in Copper Is Explained by Just Three Factors: Grade, Global GDP Growth and LME Inventory 11

9 Our Analysis Gives Rise to the Following Trend Line for China's Copper Intensity 12

10 Copper Makes the Second-Most Important Contribution to the Cash Generation of the Miners 13

11 Glencore Xstrata Is the Most Highly Exposed in Our Coverage to the Copper Price and Vale Is the Least 13

12 Copper Is the Second-Most Important Contributor to the Cash Generation of the Miners… 16

13 …And Has Been Consistently Strong Over the "Super-Cycle" 16

14 Glencore Xstrata Is the Most Highly Exposed in Our Coverage to the Copper Price While Vale Is the Least 16

15 There Is a Very Strong Relationship Between Geological Abundance and Industrial Use; the Most Useful Commodities, in Terms of Economic Application, Also Happen to Be the Most Geologically Abundant 17

16 Copper Stands Out as the Most Overutilized Commodity Relative to Its Underlying Geological Endowment — a Testimony to the Industrial Importance of This Metal and the Difficulty in Growing Supply in Anything Other Than a Supportive Price Environment 17

17 Chile Is Far More Important to Global Copper Supply Than Australia Is to Global Iron Ore; However, in Both Commodities, the Production from China (Despite the Poverty of Geological Endowment) Is Highly Significant 18



18 2004-06 Saw a "Fly-Up" in Commodity Prices, Which, When Compared to the Previous Cycles, Look Anomalous; We Strongly Believe That the Same Microeconomic Forces Are at Work Now as Previously and That Fundamentals, Rather Than "Funds," Offer the Best Explanation for the Movements in Commodity Prices 19

19 At First Glance, the Old Heuristic of Marginal Cash Costs Appeared to Have Broken Down… 20

20 …But a Better Answer Is Provided by the Negligible Visibility Into China's Mining Costs 20

21 Virtually Nothing Is Known About the Cost Structure of Chinese Copper Mining (as Shown in the Previous Exhibit), Yet the Relationship Between Price and Chinese Copper Output Tells Us That They Are Not Unrelated Phenomena 21

22 The Apparent Cost Structure of the Copper Industry in China Is Based on a Highly Non-Representative Selection of Mines; Moreover, It Fails to Adequately Describe the Profitability of This Sector 22

23 A Corrected Chinese Cost Curve Would Display Much More High-Cost Production 22

24 Calculating the Volume and Costs of the Missing Portion of the Chinese Data Is Possible Using Data from the NBS 22

25 This Mined Tonnage Needs to Be Factored Into the Global Cost Curve... 23

26 ...Where It Is Concentrated Towards Right-Hand Side of the Curve, Filling Out the High-Cost Volume 23

27 Chinese Mining Employment Started to Increase Slowly Post 2000... 24

28 ...With Mining Wages Tracking Exponential Growth Seen in the Economy as a Whole 24

29 In USD Terms, Labor Costs Have Grown in Mining at a CAGR of 28% Since 2003… 24

30 …Which Is Significantly Above the Growth Rates in Domestic Mining Output 24

31 Chinese Mining Productivity Started to Decelerate from 2003… 25

32 …And a Fundamental Disconnect Appeared Between Mining Wages and Output 25

33 The Exponential Rise in Chinese Mining Labor Costs Has Driven a Corresponding Rise in Commodity Prices 25

34 If the Cyclical Impact of Global GDP Is Added, an R-Squared of 96% Is Returned 25

35 Coal Mining in the U.K. in 1840… 27

36 …And in China Today 27

37 Nine Elements Account for 99% of the Earth's Mass; Iron and Aluminum Are Among Them, While Copper's Abundance Is Radically Different 30

38 A Very Strong Relationship Exists Between a Commodity's Geological Abundance and Its Industrial Use; the Most Useful Commodities in Terms of Economic Application Also Happen to Be the Most Geologically Abundant 31



39 Copper Stands Out as the Most Overutilized Commodity Relative to Its Underlying Geological Endowment; This Testifies to the Industrial Importance of This Metal and the Difficulty in Growing Supply in Anything Other Than a Supportive Price Environment 31

40 Below Copper Grade of 0.1%, a Step-Change Emerges in the Cost Structure of Copper Extraction; Consequently, We Face a Hard Stop in Our Ability to Exploit This Metal, Once High-Grade Deposits Are Exhausted 32

41 Under a Unimodal Distribution, Grade and Tonnage Are Continuous 33

42 Under a Bimodal Distribution, Grade and Tonnage Are Discontinuous, With Clear Implications for Cost and Availability of New Material 33

43 Copper Ores Represent the Dissemination of High-Grade Copper-Bearing Minerals Within a Barren Matrix; Copper Mining Is the Process That Separates These Valuable Minerals from the Worthless Gangue 33

44 Copper Mining Involves the Identification, Liberation and Sale of Copper-Bearing Minerals; This Is Achieved Through Two Processes of Waste Removal; the First One Occurs at the Mine Site Where Ore Is Separated from Waste; the Second One Happens at the Milling/Flotation Site Where Concentrate Is Separated from Tailings 34

45 The Traditional Mining Route, Involving the Concentration of Sulphide Ores, Drives the Vast Majority of Mined Copper Production (~80%) 37

46 The SxEw Route Exploits Oxide Ores and Accounts for the Residual 20% of the Mined Copper Production 37

47 An Important Feature of the Sulphide Route Is the Ability to Take Advantage of High-Grade Copper Ores in Secondary or Supergene Enrichment Zones; These Locations Can Provide Significant Additional Early Stage Cash Flows for a Miner 38

48 In the Distribution of Known Copper Deposits, the Graph Shows the Typical Inverse Relationship Between Grade and Size, With Copper Porphyry Occupying a Place of Privilege at One End of That Distribution 41

49 Copper Porphyries Represent the Lowest Grade But the Most Abundant Source of Copper Supply; Their Exploitation by Bulk Mining Methods Enabled the Development of the Modern Power-Intensive Industrial Society 42

50 The Average In Situ Geological Abundance in Copper Porphyry Deposits Is 0.5%, Which Has Important Implications for the Long-Term Grade Profile of Copper Production; Anything Higher Than This Must Be Temporary 42

51 The World's Consumption of Copper Is Predicated Upon the Exploitation of the Lowest-Grade Copper Deposits 43

52 Poland, Zambia and the DRC Are High Grade But Too Small to Displace American Preeminence 43

53 Chile's Dominance in the Ability to Supply the World's Copper Demand Is Clear 44

54 Chile Is Not Only Uniquely Well-Endowed in Absolute Tons, But the Size of Its Deposits Is Unmatched 44



55 Chile Stands Out as the Most Important Source of Copper Supply; However, This Was Not Always the Case, and the Study of Chile's Development Is Critical to an Understanding of the Future Copper Price Trajectory 45

56 While Chile's Growth Has Slowed Down Over the Last Decade, China's Has Accelerated 46

57 It Was the 20 Years from 1980 to 2000 That Saw Chilean Copper Growth Explode 46

58 However, Since 2000, Chile Has Stagnated and China Emerged as the Second Largest Producer of Mined Copper 46

59 Peru's Growth Has Been Sporadic 47

60 The U.S. Has Played a Critical Role in the Copper Industry and Was, for a Long While, the World's Largest Producer 47

61 As With Chile, Australian Output Has Reached a Plateau 47

62 Meanwhile, Zambia Has Only Just Recovered from the Effects of the Previous Nationalization 47

63 The Same Stagnation That We See in Chile (Once the Limits of Geology Are Reached) Is Evident in the CIS 48

64 The Impact of the Great Lakes Conflict on DRC Output Is Painfully Clear 48

65 Canada, While Still a Significant Producer, Is in Decline 48

66 Mexico Is Likely to Become a More Important Producer Over Time 48

67 In 1900, the U.S. Played a Very Similar Role in the Global Copper Supply as Chile Does Today; the Development of the Supply Side of the Copper Industry Over the Last Century Is the History of the Volume Transition from the U.S. to Chile 49

68 By 1925, the Importance of Chile Was Starting to Become Clear... 50

69 ...Though Africa (Zambia and DRC) Have Always Had a Role to Play in the Supply of Copper 50

70 By 1975, the Dominance of the U.S. in the Supply of Copper Was Already Challenged 51

71 However, It Was the Market Reforms That Inaugurated the "Miracle of Chile" That Saw the U.S. Finally Topple as the Superior Geology and Lower Labor Costs Established Chile as the Leading Global Copper Producer 51

72 Despite All of These Changes, Copper Production Remains Highly Concentrated in Very Few Regions 52

73 An Early Steam Shovel at Mt. Morgan Copper Mine in Australia at the Turn of the 20th Century 52

74 An Electric Rope Shovel at the Turn of the 21st Century — Same Idea, Slightly Bigger Scale 53

75 The Long-Term History of Copper Porphyry Discoveries Shows the Marked Impact That These Deposits Have Had on the Real Copper Price; Structurally Falling Prices Have Been Associated With Increased Finds of Relatively Few Massive Ore Bodies; We Are Not Currently in Such a Situation — There Is No New Chile on the Horizon 53

76 If We Chart Countries' Current Copper Supply vs. Their Underlying Endowment, We See an Understandably Strong Relationship; Good Geology Tends to Imply Easy Mining 55



77 It Is the Departures from This Relationship That Are Interesting; Chile Is No Longer an Easy Win and the Largest Source of Mined Copper Growth; Over the Last Decade, China Has Been Producing at More Than Twice the Level Its Geology Would Suggest 56

78 The History of Chilean Copper Development Is One of Massive Increases in Resources... 58

79 ...Occasioned by the Ability to Exploit Ever Lower-Grade Material 58

80 However, the Net Result Is a Massive Increase in Available Metal... 58

81 ...And an Even Greater Rise in Metal Output; Hence, There Has to Be More to Chile Than Increasing Discovery Rates... 58

82 ...And There Is — It Is the Increased Efficiency in Exploiting Existing Material as Seen in Mine Life Reductions 59

83 A Highly Non-Linear Relationship Exists in the Value Proposition Represented by Mine Life Reductions; They Represent Efficiency Gains Down Only to ~30 Years LOM; Afterwards, They Become Value Destructive 59

84 For a Given Geology, Halving a Mine's Life from 50 Years to 25 Years Is a Highly Profitable Exercise 60

85 Halving a Mine's Life from 30 Years to 15 Years (from the Same Geology) Is a Value Destructive Exercise 60

86 This Goes Some Way to Explaining Why Chilean Metal Output Stalled After Hitting 5.5Mtpa and a 35-Year Average LOM; It Also Highlights Why Chinese Production Looks Challenged and Where the "Low(ish) Hanging Fruit" Lie 61

87 Mine Life Expansions from Existing Reserves Have the Potential to Deliver Less Than One-Third of the Required Copper Demand by 2030; Projects Exploiting New Resources Will Be Needed and This Requires New Finds of Copper 61

88 However, the Exploitation Is Looking Ever Less Likely to Yield Results; Even If Massive New Finds Are Encountered, the History of Pebble and Oyu Tolgoi Tells Us That It Will Take 20-30 Years for These Finds to Deliver Commercially Meaningful Metal (Compared to 10 Years for Escondida) 62

89 We Look at Project-Size Differentiation, i.e., the Difference Between Percentage of Mines and Supply 63

90 This Measure Shows a Striking Correlation to Commodity Returns 63

91 Against a Fully Loaded Cost of Capital, Many of Today's New Projects Will Struggle to Create Value 63

92 The Majority of the World's Largest Projects Are in the Hands of Relatively Few Major Mining Houses 63

93 Unlike With Iron Ore, There Are Some Genuine Supply Side Constraints That Have Prevented Chilean Copper Supply from Responding to Price Increases in the Same Way as Australian Iron Ore Did 65

94 In Mining, Grade Is King; It Acts as Geological Gearing of the Labor Cost; Halve the Grade and You Will Double (at Least) the Unit Cost of Metal Output 65

95 We Need to Be Careful When Referring to Declines and Understand How Those Declines Are Measured 68



96 The Cut-Off Grade Delineates the Boundary Between the Material That Will Be Mined and Treated and the Waste Material That Will Need to Be Mined to Access the Ore, But Will Be Dumped Subsequently 69

97 For Purposes of Illustration, We Assume a Grade Profile That Declines Exponentially With Depth; at Some Depth, Grade Will Be Insufficient to Allow the Material to Be Mined Economically 70

98 The Total Tonnage That Will Be Classified as Ore Depends on How Much the Cut-Off Grade Is Allowed to Fall; the Lower the Grade That a Miner Is Prepared to Exploit, the More Tons Are Available 71

99 The Average Grade of the Tons Mined Will Naturally Be Higher With a Higher Cut-Off Grade, as There Will Be Less Low-Cost Material Present to Dilute the High-Grade Ore 71

100 However, the Relationship Between the Average Reserve Grade and the Cut-Off Grade Is Non-Linear; as Reserve Grade Falls, the Miner Is Increasingly "Scraping the Bottom of the Barrel" 72

101 The Mine Plan Determines the Sequencing of Blocks of Material Extracted from the Reserve Base; in This Instance, the Block Taken and the Block Left Behind Are Identical from a Grade Perspective; as a Result, the Mine Plan Sees a Grade Profile That Is Constant Over Time 73

102 Under a Different Sequence, the Mine Plan Removes the Highest Grade First, Leaving Lower Grade Material Behind; Over Time, This Means That the Head Grade Must Fall as the Cut-Off Grade Limit Is Approached 74

103 Three Different Grade Profiles for the Development of Exactly the Same Underlying Ore Body... 76

104 ...Lead to Radically Different Value Propositions; Under Realistic Mining Investment Guidelines, Only the Third Scenario of Declining Ore Grade Will Get Approved and Subsequently Developed; the Other Two Mining Solutions Are Unlikely to Attract Capital 77

105 While Generating the Same Total Lifetime Cash, High-Grading the Ore Body Delivers Cash Sooner 77

106 Geology Conspires With Finance, as Secondary Enrichment Zones Help Create High-Grade Ore Near the Surface, Which Can Be Extracted Early on in the Mine Life 78

107 The Basic Structure of the Operating Costs of a Stylized Large Chilean Copper Mine 80

108 The Picture for a Small Chinese Type Operation (or at Least What We Believe They Look Like!) Is a Little Different 81

109 Stripping Ratio: The Deeper You Go the Higher the Cost 82

110 Labor Productivity: Efficient and Well-Capitalized Mines Are Far Lower Cost Than Those Relying on Human Labor 82

111 The Price of Labor Matters... 82

112 ...While the Impact of Diesel Is Surprisingly Low 82

113 Power Has Less Impact Than One Might Expect 83

114 Rock Hardness Affects the Milling Power Rate and Drives Cost in High Power-Cost Regions 83

115 Higher Mill Recovery (Finer Grind) Implies Lower Cost 83

116 Nevertheless, in All of This "Grade Is King" 83



117 Compounding the Critical Role That Grade Plays Is the Fact That the Productivity of Milling Circuits Varies With It; a "Constant Tail" of Copper Is Always Lost, Given That Below a Certain Copper Concentration, Froth Flotation Cannot Discriminate Between Ore and Tailings 84

118 This Radically Increases the Cost of Low-Grade Volume Exploitation 84

119 The Life of Available Copper Deposits Has Held Up Very Well, Despite the Massive Increases in Mined Production 85

120 So There Is No Need for Any Malthusian Concerns; We Have as Much Copper "on Hand" Now as 35 Years Ago 85

121 This Security Is Engendered by the Fact That Reserves Seem to Have Held Up Well 86

122 However, the Truth Is Somewhat More Complicated; Reserves Have Held Up Well Only Because of a Dramatic Reduction in the Reserve Grade... 86

123 ...Driven by Dropping the Cut-Off Grade to below 0.3% Cu 86

124 While the Decline in Head Grade Is Well Known... 87

125 ...Its Implications Are Not, Given the Nature of the Copper Forward Curve and Consensus Price Expectations 87

126 Production Over the Last Decade Has Been Sustained Only Through Moving Head Grades Above Reserve Grade and Global High-Grading 87

127 This Means That Grades Must Decline Sharply in the Future; in Fact, Grades Must Decline Towards the Much Lower Level Given by the Cut-Off Grade 88

128 2003 (or Thereabouts) Saw a Structural Break in the Industry, Which Moved from a Period Where Mining Did Not Imply Geological Degradation to One Where It Most Certainly Does 88

129 Just for Fun, We Provide a Long-Term History of Copper Grades in Australia 89

130 But on a Serious Point: Post-2000, We See Stagnation in the Copper Industry, Which Saw Mine Output Holding Up Only Through a Radical Grade Decline; What Happened in Australia Was Repeated Globally 90

131 The Relationship Between Cut-Off Grade, Head Grade and Reserve Grade Over the Life of the Reserve Base Enables One to Calculate How Head Grades Will Need to Evolve Going Forward 92

132 How Is All This Related to Price? Well, in the First Instance, We Can Observe That the Step-Up in Prices Occurred When Global Head Grades Started to Fall 93

133 Indeed, There Is a Very Strong Relationship Between These Two Factors 93

134 We Use This as the Basis of a Simple Multivariate Regression, Which Almost Perfectly Describes the Copper Price 94

135 Nearly 90% of the Movement in Copper Is Given by Just Three Factors: Grade, Global GDP Growth and LME Inventory 94

136 The Normalcy of the Errors Between Regressed and Actual Copper Prices Implies No Systemic Bias or Neglected Explanatory Variable… 95

137 …Leading to a Regression Analysis Whose Component Parts Are All Highly Significant 95



138 Grade Declines Are An Irreversible Feature of the Industry, Driven by the Structure of Current Mine Plans That Stand Behind Current Copper Consumption 96

139 Grade Is by Far the Most Important to Price Sensitivity… 96

140 …But Economic Growth Prospects Can Swing the Price by ~US1,500/t… 97

141 …With a Similar Order of Magnitude Contribution from Terminal Market Inventory 97

142 This Implies That GDP Has to "Work" Incredibly Hard to Overcome the Inevitable Grade Decline and Return Consensus Price Expectations; to Put It Mildly, the Implicit Assumption of Consensus Is Implausible… 98

143 …Likewise for Terminal Market Inventory; Surely, the Most Likely Read Through Is Not That the Miners Forget How to Run Their Businesses, But That Prices Will Trend Higher Than Consensus Currently Anticipates 98

144 Grade Is by Far the Strongest Driver of Price; Under Any Scenario, Dropping the Cut-Off Grade to the Point Where the Head Grade Would Be Below the Reserve Grade Implies Prices Well in Excess of US$10,000/t 99

145 Labor (Including Services) and Consumables Represent the Largest Cost Elements for the Mine Site 103

146 Consumables and Power Form the Largest Cost Element for Milling 103

147 Overhead Costs of a Mine Site Mainly Comprise Labor 103

148 The Split Between Mining and Milling Is Roughly Even With G&A Being a Relatively Small Part of the Overall Costs 103

149 The Price of Diesel Shows Relatively Small Variation from One Mining Jurisdiction to the Next 104

150 Chile Stands Out as a Very High Power Cost Region; in Africa, It Is Not the Cost But Rather the Reliability of Power Supply That Is the Issue 104

151 Unsurprisingly, the Developed World Has the Highest Labor Cost; Cheap Labor (as Well as High Grades) Will Drive the Value of the African Copper Production; Chile Is Already a Medium- to High-Cost Mining Location 105

152 While Diesel and Power Are Tied to Global Pricing Trends, Labor Shows the Highest Variation Across Regions 105

153 A Belief in Declining Labor Costs Is at Odds With the Reality of Demands for Increasing Nominal Wages 106

154 Power Costs Are Historically Highly Correlated With Oil Prices... 106

155 ...And a Belief in an Immediate and Sharp Downward Correction Is Not the View We Endorse 106

156 The Aggregate Cost Escalator That Stands Behind the Consensus View of the Copper Price Seems, at Least to Us, Highly Implausible 106

157 Nominal Wage Increases and the Expectation of a Higher Living Standard Over Time Form the Basis of the Political Mandate in Most Developed and Developing Countries 107

158 We Take a Different View from Wood Mackenzie (and Hence from Consensus) on How Mining Costs Will Evolve, With the Differences Being Particularly Pronounced for Expected Future Labor Costs 108



159 Our Base Case Has Continued Appreciation in the RMB... 108

160 ...And Significant Growth in Nominal RMB-Denominated Output 108

161 At the Same Time, We Expect the Chinese Workforce to Continue to Decline 109

162 Overall, We Expect the Dollar Output per Chinese Worker to Increase Significantly, and This Drives Our Forecast for the Chinese Mining Labor Cost Escalation 109

163 Added to This Is the SCB View on the Evolution of the Oil Price... 110

164 ...Which We Believe Will Imply a Continuation of the Increase in the Nominal Price of Energy 110

165 Adding These Factors Together Gives a Country-Specific Mining Cost Escalator... 110

166 ...As Well as Milling Cost... 110

167 ...And Ancillary Costs... 111

168 ...For a Total Cost Escalator for Mined Copper on a Region-by-Region Basis 111

169 It Is the Differential Cost Escalation That Drives Real Commodity Price Increases in USD Terms 112

170 The Increase in Cost Pressure Is Powerfully Attested to by the Increase in Ore That Needs to Be Treated... 112

171 ...And the Declining Grade of the Ore That Is Processed 112

172 It Implies Lower Mill Recoveries 113

173 All These Factors Feed Into Our Expectation of Decreasing Labor Productivity and a Macro Environment With Ever-Increasing Nominal Wages 113

174 Against a Fully Loaded Cost of Capital, It Is Clear That Many of Today's New Projects Will Struggle to Create Value 114

175 There Is Always a Bias to Execute the Best Projects First; Even So, a Real Price of Above US$9,000/t Will Be Required to Bring the New Projects On Line Over the Next Decade 115

176 A Key Component of the Required Price Is Due to the Requirement to Compensate for the Risk and Losses That Will Be Borne as a Consequence of More Supply Coming from New Frontier Regions 116

177 The Current World Production of Copper Has Been Achieved Through Targeting the Easy Win Locations, Which Combined High Geological Prospectivity With Low Political Risk; This Model No Longer Holds and Higher Prices Will Be Required to Reflect This Reality 117

178 We Believe That Low-Grade High-Cost Chinese Producers Already Sit at the Right-Hand Side of the Global Cost Curve and That Their Presence There Is Poorly Understood; As Real USD Cost Escalation in China on the Back of Increasing Labor Costs Takes Hold, It Will Steepen the Cost Curve and Lead to Ever Higher Copper Prices 119

179 Post the Oil Shocks of the 1970s, There Was a Step-Change in Demand for Copper; While Growth Slowed, It Never Went Ex-Growth in a Trend Sense 122

180 Global Copper Growth Is Still Not as Strong as It Was Post World War II... 122

181 ...With the West Acting as a Drag on Overall Demand 122



182 China and Developing Asia Have Taken Over the Lead in Copper Demand 123

183 The Current Cycle Shows Markedly Lower Copper Growth Than Steel 123

184 How Much of the First Use Copper Demand Decline Is Attributable to Imports of Metal in Other Forms? 124

185 We Believe That Only About 50% of Copper Use Is "Exportable" 124

186 In Our View, Net Export to Import Positions Are Representative of the Overall Flow of Copper in Goods 124

187 This Yields the Following "Finished to Final" Use Correction 124

188 Japan Presents an Interesting Precedent for Current Chinese Copper Consumption 126

189 The Copper Consumption in Japan Can Best Be Understood by Looking at Copper Intensity 126

190 However, There Needs to Be a Way of Bridging Between Metal Intensity in One Time Period to Another; We Find That Bridge in Metal (Capital) Stock Formation 128

191 Unsurprisingly, a Very Strong Relationship Exists Between the Overall Level of Capital Stock in a Country and Its Output; We Link the Metal Intensity in Different Time Periods Through the Assumption That the Rate of Overall Economic Development Must Be Accompanied by a Corresponding Development in Capital Stock 129

192 In Order to Forecast China's (as Well as Any Other Country's) Consumption, We Begin With the History of Copper Consumption... 130

193 ...And the Corresponding Copper Intensity... 130

194 …That Gives Rise to a Corresponding Development in Capital Stock 131

195 We Expect the Chinese Copper Stock Development to Resemble That of Japan; We Use This Expectation to Calculate Both How Copper Intensity in Any Developing Country Tracks the Overall Economic Development and the Multiplier Between Metal Growth and Economic Growth at Any Point in Time 131

196 This Gives Rise to the Following Trend Line for China's Copper Intensity... 132

197 ...And a Corresponding Development in Its Copper Consumption; We Expect It to Peak at 14Mtpa Post-2025; in This Regard, Copper Is a "Later Cycle" Commodity Than Steel, Whose Peak We Anticipate Nearer 2020 132

198 We Then Aggregate Each of the Country-Specific Demand Forecasts to Arrive at a Global Total That Enables Us to Derive a Picture of Still Rising Copper Intensity Out Till 2020 133

199 Nonetheless, China Remains the Most Important Driver of Overall Demand 133

200 Labor (Including Services) and Consumables (Tires, etc.) Represent the Largest Cost Elements for the Mine Site... 136

201 ...While It Is Consumables (Grinding Media and Reagents) and Power That Form the Largest Cost Element for Milling 136

202 The Variation in Labor Costs Is the Most Important Source of Differentiation for Cost Escalators 136



203 Adding in Currency Effects Generates the Overall Cost Escalator for Each Region 136

204 The Increase in Cost Pressure Is Powerfully Attested to by the Increase in Ore That Needs to Be Treated... 137

205 ...And the Declining Grade of the Ore That Is Processed 137

206 It Implies Lower Mill Recoveries 137

207 All These Factors Feed Into Our Expectation of Decreasing Labor Productivity and a Macro Environment With Ever Increasing Nominal Wages 137

208 In Aggregate, We See a Balanced to Slightly Oversupplied Copper Market for the Next 18 to 24 Months (Assuming No Further Supply Side Shocks or Demand-Side Weakness); However, Post This Period, Current Capacity Falls Well Short of Demand 138

209 There Is a Sufficient "Reservoir" of New Projects to Fill Any Supply/demand Gap; the Only Question Is the Price Required to Incentivize These Projects to Come on Line 140

210 The Current Low Price for Copper (and Copper By-Products Such as Gold and Molybdenum) Makes It Very Hard for New Projects to Get Approved, Thus Opening up a Supply Gap in the 2015E-20E Period 140

211 We See the U.S. as One of the Biggest Winners (Along With Peru) in the Supply of Future Copper 141

212 However, African Growth Is an Absolute Requirement for Longer-Term Copper Supply 141

213 We Are Getting Increasingly Worried About the Longer-Term Ability of Chile to Maintain Its Capacity 142

214 We Expect Copper Price to Remain High for as Long as China's Wage Inflation Grows Faster Than Its Mining Productivity, Unless the Western Miners Expend So Much Capital as to Displace the Need for Such Marginal Activity Entirely 143

215 We See a Supply-Demand Balance in the Immediate Short Term; We Expect the Market to Become Increasingly Stretched Over the Next Two Years; by Then We Expect Cost Escalation and Productivity in China (and Elsewhere) to Have Only Deteriorated Further, Thus Leading to Ever Higher Prices Necessary to Incentivize the Next Wave of Mine Investment 144

216 1.5% of Mines Account for 25% of Global Copper Supply 145

217 Understanding the Development of These 10 Mines Is Critical in Understanding the Forward-Looking Copper Balance 145

218 3% of New Projects Account for 18% of Possible New Supply 146

219 It Is Rio Tinto and Glencore Xstrata That Have the Greatest Exposure to These Developments 146

220 Escondida Overview 147

221 Escondida Ownership 147

222 Escondida Metal Exposure 147

223 Escondida Geological Endowment 148

224 Escondida Ore Grade 148

225 Escondida Head Grade 148

226 Escondida Cost Position 148

227 Antamina Overview 149

228 Antamina Ownership 149



229 Antamina Metal Exposure 149

230 Antamina Geological Endowment 150

231 Antamina Ore Grade 150

232 Antamina Head Grade 150

233 Antamina Cost Position 150

234 Los Pelambres Overview 151

235 Los Pelambres Ownership 151

236 Los Pelambres Metal Exposure 151

237 Los Pelambres Geological Endowment 152

238 Lost Pelambres Ore Grade 152

239 Los Pelambres Head Grade 152

240 Los Pelambres Cost Position 152

241 El Teniente Overview 153

242 El Teniente Ownership 153

243 El Teniente Metal Exposure 153

244 El Teniente Geological Endowment 154

245 El Teniente Ore Grade 154

246 El Teniente Head Grade 154

247 El Teniente Cost Position 154

248 Chuquicamata Overview 155

249 Chuquicamata Ownership 155

250 Chuquicamata Metal Exposure 155

251 Chuquicamata Geological Endowment 156

252 Chuquicamata Ore Grade 156

253 Chuquicamata Head Grade 156

254 Chuquicamata Cost Position 156

255 Grasberg Overview 157

256 Grasberg Ownership 157

257 Grasberg Metal Exposure 157

258 Grasberg Geological Endowment 158

259 Grasberg Ore Grade 158

260 Grasberg Head Grade 158

261 Grasberg Cost Position 158

262 Los Bronces Overview 159

263 Los Bronces Ownership 159

264 Los Bronces Metal Exposure 159

265 Los Bronces Geological Endowment 160

266 Los Bronces Ore Grade 160

267 Los Bronces Head Grade 160

268 Los Bronces Cost Position 160

269 Radomiro Tomic Overview 161

270 Radomiro Tomic Ownership 161

271 Radomiro Tomic Metal Exposure 161

272 Radomiro Tomic Geological Endowment 162

273 Radomiro Tomic Ore Grade 162



274 Radomiro Tomic Leach Grade 162

275 Radomiro Tomic Cost Position 162

276 Andina Overview 163

277 Andina Ownership 163

278 Andina Metal Exposure 163

279 Andina Geological Endowment 164

280 Andina Ore Grade 164

281 Andina Head Grade 164

282 Andina Cost Position 164

283 Collahuasi Overview 165

284 Collahuasi Ownership 165

285 Collahuasi Metal Exposure 165

286 Collahuasi Geological Endowment 166

287 Collahuasi Ore Grade 166

288 Collahuasi Head Grade 166

289 Collahuasi Cost Position 166

290 Oyu Tolgoi Overview 167

291 Oyu Tolgoi Ownership 167

292 Oyu Tolgoi Metal Exposure 167

293 Oyu Tolgoi Grade Profile 168

294 Oyu Tolgoi Production Profile 168

295 Oyu Tolgoi Geological Endowment 168

296 Oyu Tolgoi Grade 168

297 Resolution Overview 169

298 Resolution Ownership 169

299 Resolution Metal Exposure 169

300 Resolution Grade Profile 170

301 Resolution Production Profile 170

302 Resolution Geological Endowment 170

303 Resolution Grade 170

304 La Granja Overview 171

305 La Granja Ownership 171

306 La Granja Metal Exposure 171

307 La Granja Grade Profile 172

308 La Granja Production Profile 172

309 La Granja Geological Endowment 172

310 La Granja Grade 172

311 Tampakan Overview 173

312 Tampakan Ownership 173

313 Tampakan Metal Exposure 173

314 Tampakan Grade Profile 174

315 Tampakan Production Profile 174

316 Tampakan Geological Endowment 174

317 Tampakan Grade 174

318 Escondida Overview 175



319 Escondida Ownership 175

320 Escondida Metal Exposure 175

321 Escondida Grade Profile 176

322 Escondida Production Profile 176

323 Escondida Geological Endowment 176

324 Escondida Grade 176

325 Golpu Overview 177

326 Golpu Ownership 177

327 Golpu Metal Exposure 177

328 Golpu Grade Profile 178

329 Golpu Production Profile 178

330 Golpu Geological Endowment 178

331 Golpu Grade 178

332 Las Bambas Overview 179

333 Las Bambas Ownership 179

334 Las Bambas Metal Exposure 179

335 Las Bambas Grade Profile 180

336 Las Bambas Production Profile 180

337 Las Bambas Geological Endowment 180

338 Las Bambas Grade 180

339 Los Pelambres Overview 181

340 Los Pelambres Ownership 181

341 Los Pelambres Metal Exposure 181

342 Los Pelambres Grade Profile 182

343 Los Pelambres Production Profile 182

344 Los Pelambres Geological Endowment 182

345 Los Pelambres Grade 182

346 Pebble Overview 183

347 Pebble Ownership 183

348 Pebble Metal Exposure 183

349 Pebble Grade Profile 184

350 Pebble Production Profile 184

351 Pebble Geological Endowment 184

352 Pebble Grade 184

353 Andina Overview 185

354 Andina Ownership 185

355 Andina Metal Exposure 185

356 Andina Grade Profile 186

357 Andina Production Profile 186

358 Andina Geological Endowment 186

359 Andina Grade 186



Disclosure Appendix

VALUATION METHODOLOGY

As mining companies represent operationally and financially geared exposure to underlying commodity baskets (with ~80% of weekly equity price moves explained by moves in underlying commodity prices), we use a regression-based trading model and our forward commodity price forecasts to determine our 12-month price targets for our European metals and mining coverage.

To the extent that the regression holds, and the parameters of the regression have not significantly shifted, we take the target price from the trading model. To the extent that the regression is shifting or the equity is deviating, we look for evidence of whether this shift or deviation is temporary (and hence may be expected to close) or whether it signals a more fundamental re- or de-rating of the equity. In the event that there is no significant deviation or if we believe a deviation is temporary, the target price is set by the trading model. In the event that we believe a deviation is signaling a fundamental change, we will adjust our target price for this fundamental shift and disclose the manner and magnitude of the adjustment made. At present, no adjustments have been made to the target prices generated by our trading model. Note that we round final target prices in 25p/cent increments.

In addition to the target price (and short-term price forecasts generated by our trading model), we calculate a supplementary valuation that is DCF based. Given the long-lived nature of mining assets, we believe a DCF is critical to understanding the intrinsic value of a share (what the share price, in our view, "ought" to be today). Our DCF model is constructed in nominal local currency terms out to 2030, over which explicit commodity price and exchange rate forecasts apply. The nominal local currency cash flows are de-escalated into real USD cash flows and discounted at the company-specific WACC. A country risk premium reflecting the geographic origin of the cash flows is added to the underlying WACC to reflect cash flow items (i.e., expropriation) that cannot be explicitly modeled in the cash flow. All reserves are considered exploited by the model. In addition 50% of the incremental resources (i.e., 50% of the residual resources, excluding those that have already been converted to reserves) of the company are modeled. Where residual life of the mine (LOM) may be inferred for operations beyond the 2030 time horizon, a terminal value is calculated for the remaining years of potentially exploitable material. We use this methodology to derive all forward-looking multiples and other valuation metrics. Note that we forecast our models in reporting currency (USD), convert to listing currency (British pound sterling or Brazilian real) at an average exchange rate, and round final DCF values in 25p/cent increments.

RISKS

The four most significant risks facing the major mining houses are: 1) lack of capital discipline (specifically displacement of high-cost Chinese marginal producers by low-cost Western production), 2) operating cost inflation (USD-denominated unit costs in all the major mining houses have seen double-digit growth rates over the last 10 years, roughly half of which are macro related and the other half are real local currency), 3) a sustained downturn in the Chinese economy (the largest consumer of global resources) and 4) resource nationalism (ranging from increased share of rent extraction to outright asset confiscation).

SRO REQUIRED DISCLOSURES

References to "Bernstein" relate to Sanford C. Bernstein & Co., LLC, Sanford C. Bernstein Limited, Sanford C. Bernstein (Hong Kong) Limited, and Sanford C. Bernstein (business registration number 53193989L), a unit of AllianceBernstein (Singapore) Ltd. which is a licensed entity under the Securities and Futures Act and registered with Company Registration No. 199703364C, collectively.

Bernstein analysts are compensated based on aggregate contributions to the research franchise as measured by account penetration, productivity and proactivity of investment ideas. No analysts are compensated based on performance in, or contributions to, generating investment banking revenues.

Bernstein rates stocks based on forecasts of relative performance for the next 6-12 months versus the S&P 500 for stocks listed on the U.S. and Canadian exchanges, versus the MSCI Pan Europe Index for stocks listed on the European exchanges (except for Russian companies), versus the MSCI Emerging Markets Index for Russian companies and stocks listed on emerging markets exchanges outside of the Asia Pacific region, and versus the MSCI Asia Pacific ex-Japan Index for stocks listed on the Asian (ex-Japan) exchanges - unless otherwise specified. We have three categories of ratings:

Outperform: Stock will outpace the market index by more than 15 pp in the year ahead.

Market-Perform: Stock will perform in line with the market index to within +/-15 pp in the year ahead.

Underperform: Stock will trail the performance of the market index by more than 15 pp in the year ahead.

Not Rated: The stock Rating, Target Price and estimates (if any) have been suspended temporarily.

As of 09/24/2013, Bernstein's ratings were distributed as follows: Outperform - 41.3% (0.9% banking clients) ; Market-Perform - 46.2% (0.4% banking clients); Underperform - 12.5% (0.0% banking clients); Not Rated - 0.0% (0.0% banking clients). The numbers in parentheses represent the percentage of companies in each category to whom Bernstein provided investment banking services within the last twelve (12) months.

Accounts over which Bernstein and/or their affiliates exercise investment discretion own more than 1% of the outstanding common stock of the following companies RIO.LN / Rio Tinto PLC, BLT.LN / BHP Billiton PLC.

This research publication covers six or more companies. For price chart disclosures, please visit www.bernsteinresearch.com, you can also write to either: Sanford C. Bernstein & Co. LLC, Director of Compliance, 1345 Avenue of the Americas, New York, N.Y. 10105 or Sanford C. Bernstein Limited, Director of Compliance, 50 Berkeley Street, London W1J 8SB, United Kingdom; or Sanford C. Bernstein (Hong Kong) Limited, Director of Compliance, Suites 3206-11, 32/F, One International Finance Centre, 1 Harbour View Street, Central, Hong Kong, or Sanford C. Bernstein (business registration number 53193989L) , a unit of AllianceBernstein



(Singapore) Ltd. which is a licensed entity under the Securities and Futures Act and registered with Company Registration No. 199703364C, Director of Compliance, 30 Cecil Street, #28-08 Prudential Tower, Singapore 049712.

12-Month Rating History as of 09/23/2013

Ticker Rating Changes

AAL.LN O (IC) 09/05/12

BHP O (IC) 09/05/12

BHP.AU O (IC) 09/26/12

BLT.LN O (IC) 09/05/12

GLEN.LN O (RC) 02/13/13 M (IC) 09/05/12

RIO O (IC) 09/05/12

RIO.LN O (IC) 09/05/12

VALE O (RC) 06/07/13 U (IC) 09/05/12

VALE3.BZ O (RC) 06/07/13 U (IC) 09/05/12

Rating Guide: O - Outperform, M - Market-Perform, U - Underperform, N - Not Rated

Rating Actions: IC - Initiated Coverage, DC - Dropped Coverage, RC - Rating Change

OTHER DISCLOSURES

A price movement of a security which may be temporary will not necessarily trigger a recommendation change. Bernstein will advise as and when coverage of securities commences and ceases. Bernstein has no policy or standard as to the frequency of any updates or changes to its coverage policies. Although the definition and application of these methods are based on generally accepted industry practices and models, please note that there is a range of reasonable variations within these models. The application of models typically depends on forecasts of a range of economic variables, which may include, but not limited to, interest rates, exchange rates, earnings, cash flows and risk factors that are subject to uncertainty and also may change over time. Any valuation is dependent upon the subjective opinion of the analysts carrying out this valuation.

This document may not be passed on to any person in the United Kingdom (i) who is a retail client (ii) unless that person or entity qualifies as an authorised person or exempt person within the meaning of section 19 of the UK Financial Services and Markets Act 2000 (the "Act"), or qualifies as a person to whom the financial promotion restriction imposed by the Act does not apply by virtue of the Financial Services and Markets Act 2000 (Financial Promotion) Order 2005, or is a person classified as an "professional client" for the purposes of the Conduct of Business Rules of the Financial Conduct Authority.

To our readers in the United States: Sanford C. Bernstein & Co., LLC is distributing this publication in the United States and accepts responsibility for its contents. Any U.S. person receiving this publication and wishing to effect securities transactions in any security discussed herein should do so only through Sanford C. Bernstein & Co., LLC.

To our readers in the United Kingdom: This publication has been issued or approved for issue in the United Kingdom by Sanford C. Bernstein Limited, authorised and regulated by the Financial Conduct Authority and located at 50 Berkeley Street, London W1J 8SB, +44 (0)20-7170-5000.

To our readers in member states of the EEA: This publication is being distributed in the EEA by Sanford C. Bernstein Limited, which is authorised and regulated in the United Kingdom by the Financial Conduct Authority and holds a passport under the Markets in Financial Instruments Directive.

To our readers in Hong Kong: This publication is being distributed in Hong Kong by Sanford C. Bernstein (Hong Kong) Limited which is licensed and regulated by the Hong Kong Securities and Futures Commission (Central Entity No. AXC846). This publication is solely for professional investors only, as defined in the Securities and Futures Ordinance (Cap. 571).

To our readers in Singapore: This publication is being distributed in Singapore by Sanford C. Bernstein, a unit of AllianceBernstein (Singapore) Ltd., only to accredited investors or institutional investors, as defined in the Securities and Futures Act (Chapter 289). Recipients in Singapore should contact AllianceBernstein (Singapore) Ltd. in respect of matters arising from, or in connection with, this publication. AllianceBernstein (Singapore) Ltd. is a licensed entity under the Securities and Futures Act and registered with Company Registration No. 199703364C. It is regulated by the Monetary Authority of Singapore and located at 30 Cecil Street, #28-08 Prudential Tower, Singapore 049712, +65-62304600. The business name "Sanford C. Bernstein" is registered under business registration number 53193989L.

To our readers in Australia: Sanford C. Bernstein & Co., LLC, Sanford C. Bernstein Limited and Sanford C. Bernstein (Hong Kong) Limited are exempt from the requirement to hold an Australian financial services licence under the Corporations Act 2001 in respect of the provision of the following financial services to wholesale clients:

providing financial product advice;

dealing in a financial product;

making a market for a financial product; and

providing a custodial or depository service.

Sanford C. Bernstein & Co., LLC., Sanford C. Bernstein Limited, Sanford C. Bernstein (Hong Kong) Limited and AllianceBernstein (Singapore) Ltd. are regulated by, respectively, the Securities and Exchange Commission under U.S. laws, by the Financial Conduct Authority under U.K. laws, by the Hong Kong Securities and Futures Commission under Hong Kong laws, and by the Monetary Authority of Singapore under Singapore laws, all of which differ from Australian laws.



One or more of the officers, directors, or employees of Sanford C. Bernstein & Co., LLC, Sanford C. Bernstein Limited, Sanford C. Bernstein (Hong Kong) Limited, Sanford C. Bernstein (business registration number 53193989L) , a unit of AllianceBernstein (Singapore) Ltd. which is a licensed entity under the Securities and Futures Act and registered with Company Registration No. 199703364C, and/or their affiliates may at any time hold, increase or decrease positions in securities of any company mentioned herein.

Bernstein or its affiliates may provide investment management or other services to the pension or profit sharing plans, or employees of any company mentioned herein, and may give advice to others as to investments in such companies. These entities may effect transactions that are similar to or different from those recommended herein.

Bernstein Research Publications are disseminated to our customers through posting on the firm's password protected website, www.bernsteinresearch.com. Additionally, Bernstein Research Publications are available through email, postal mail and commercial research portals. If you wish to alter your current distribution method, please contact your salesperson for details.

Bernstein and/or its affiliates do and seek to do business with companies covered in its research publications. As a result, investors should be aware that Bernstein and/or its affiliates may have a conflict of interest that could affect the objectivity of this publication. Investors should consider this publication as only a single factor in making their investment decisions.

This publication has been published and distributed in accordance with Bernstein's policy for management of conflicts of interest in investment research, a copy of which is available from Sanford C. Bernstein & Co., LLC, Director of Compliance, 1345 Avenue of the Americas, New York, N.Y. 10105, Sanford C. Bernstein Limited, Director of Compliance, 50 Berkeley Street, London W1J 8SB, United Kingdom, or Sanford C. Bernstein (Hong Kong) Limited, Director of Compliance, Suites 3206-11, 32/F, One International Finance Centre, 1 Harbour View Street, Central, Hong Kong, or Sanford C. Bernstein (business registration number 53193989L) , a unit of AllianceBernstein (Singapore) Ltd. which is a licensed entity under the Securities and Futures Act and registered with Company Registration No. 199703364C, Director of Compliance, 30 Cecil Street, #28-08 Prudential Tower, Singapore 049712. Additional disclosures and information regarding Bernstein's business are available on our website www.bernsteinresearch.com.

CERTIFICATIONS

I/(we), Paul Gait, Senior Analyst(s)/Analyst(s), certify that all of the views expressed in this publication accurately reflect my/(our) personal views about any and all of the subject securities or issuers and that no part of my/(our) compensation was, is, or will be, directly or indirectly, related to the specific recommendations or views in this publication.

Approved By: NK

Copyright 2013, Sanford C. Bernstein & Co., LLC, Sanford C. Bernstein Limited, Sanford C. Bernstein (Hong Kong) Limited, and AllianceBernstein (Singapore) Ltd., subsidiaries of AllianceBernstein L.P. ~1345 Avenue of the Americas ~ NY, NY 10105 ~212/756-4400. All rights reserved.

This publication is not directed to, or intended for distribution to or use by, any person or entity who is a citizen or resident of, or located in any locality, state, country or other jurisdiction where such distribution, publication, availability or use would be contrary to law or regulation or which would subject Bernstein or any of their subsidiaries or affiliates to any registration or licensing requirement within such jurisdiction. This publication is based upon public sources we believe to be reliable, but no representation is made by us that the publication is accurate or complete. We do not undertake to advise you of any change in the reported information or in the opinions herein. This publication was prepared and issued by Bernstein for distribution to eligible counterparties or professional clients. This publication is not an offer to buy or sell any security, and it does not constitute investment, legal or tax advice. The investments referred to herein may not be suitable for you. Investors must make their own investment decisions in consultation with their professional advisors in light of their specific circumstances. The value of investments may fluctuate, and investments that are denominated in foreign currencies may fluctuate in value as a result of exposure to exchange rate movements. Information about past performance of an investment is not necessarily a guide to, indicator of, or assurance of, future performance.


EUROPEAN METALS & MINING: COPPER FOR THE CRAFTSMAN CUNNING AT HIS TRADE

For the exclusive use of JA

SO

N LA

PO

RT

E at P

ER

RY

CA

PIT

AL on 25-S

ep-2013

copper market primer

Documents

t copper

copper demand

copper supplier

decades of copper

replacement of copper

copper sources of supply

rising copper price

higher copper prices