Iron ore

A new dimension

A potential adrenaline shot for the industry, SGX’s 65% Fe derivative

presents more dynamic trading opportunities for the iron ore market.

We examine how market fundamentals drive price spreads between

grades of iron ore, and analyse the extent to which sector-related
news and data may act as trading cues.
The breakneck pace of
change in the iron ore
market certainly keeps
us on our toes

Less than four years ago analysts were predicting that the continued
deterioration of global ore grades could see the SGX’s newly-launched 58%
Fe contract eventually challenge the 62% as the market’s benchmark. But
that was then. Radical changes imposed from the start of 2016 by the world’s
super-consumer have ensured the industry that recovered from the last cyclical
downturn is a very different beast to the one that went into it. The high-grade
iron ore market’s increased separation from the lower-grade segments, both in
value and correlation, has prompted SGX to launch a new derivative contract.

Cash-settled against the Fastmarkets MB 65% Fe Brazilian Fines Index – the

benchmark for the global high-grade fines market – the new derivative fulfils the
market’s need for a tool to manage price exposure to high grades ores.

The China factor

When China unveiled its 13th 5-year plan at the start of 2016 few believed that the
targets laid out to reform its steel industry were realistic or achievable. But three years
on and the country’s steel industry is all but unrecognizable to its pre-2016 look – over
150Mt of official steelmaking capacity having been cut, on top of some 140Mt of illegal,
low-quality induction furnace production. The tougher environmental protection law
that came into effect at the start of 2016 has also put mills under more stringent
operating requirements – something that’s made blue skies in Beijing no longer an
Instagram-worthy occurrence.

Despite the extensive cuts, China’s steel output – at least the ‘official’ on-the-books
production – has continued to touch record levels. Crude steel production was up
6.46% year-on-year in the January-September period of 2018 for example, as remaining
capacity operates at much higher utilization rates. With strong downstream demand,
mills have up until very recently been incentivized to raise efficiency by the handsome
profit margins available on most steel products. These higher efficiency rates have been
achieved by consuming higher-purity raw materials.

2
Factors driving inter-
grade spread volatility

Steelmaker profit margins

The profitability of Chinese steel mills over time shows a strong positive correlation with
the size of the 65-62% Fe Index differential (Figure 2). The reason is that the higher silica
and alumina levels in the 62% Index specification (Figure 1) increase blast furnace slag
volume, lowering hot metal yields and limiting steel production rates. Higher quality
ores also raise efficiency by forming stronger (lower reduction-degradation index) sinter,
which increases burden permeability and improves flow of reducing gases.

Fig. 1: Major Gangue Components: 65 vs 62% Fe Indices

The levels of iron ore’s major gangue components – silica, alumina and phosphorus – are
lower in the 65% Fe index specification than in the 62%

Phosphorus

Alumina

Silica

0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0% 3.5% 4.0%

MB65 MB62 Source: Fastmarkets MB

In a negative margin scenario (such as in 2015) lower productivity can help minimize
losses, but in a profitable market inefficiency represents an opportunity cost. Fluctuations
in inter-grade price spreads reflect this reactivity to different market conditions.

Fig. 2: Mill Margin Proxy vs Fastmarkets MB 65-62% Fe Index Spread

Proxy indicators of HRC and rebar production profit margins in China demonstrate a
clear positive correlation with the movement of the 65-62% Fe index differential. Source:
Fastmarkets MB, Steelhome (Mill margin proxies calculated using domestic prices of
rebar & HRC (East China), coke & iron ore)

1750 35
Rebar Magin Proxy (yuan/tonne) [LHS]
30
HRC Magin Proxy (yuan/tonne) [LHS]
1250 65-62% Fe Index $-Differential [RHS] 25

20

750 15

10

250 5

0
Jan-16

Jan-18
Jan-14

Oct-16

Oct-18
Jan-15
Oct-14

Oct-15

Jan-17

Oct-17
Apr-16

Apr-18
Apr-14

Jul-16

Jul-18
Apr-15
Jul-14

Jul-15

Apr-17

Jul-17

-250 -5

-10

-750 -15

3
Stricter environmental regulation

On top of the profitability incentive, Chinese mills have also been encouraged to use
higher grade ores as a response to stricter environmental policies. Not only do lower
slag volumes improve blast furnace productivity, they also lower coke consumption and
reduce pollution. As slag has a higher melting temperature than iron, extra fuel in the
form of coke is required to maintain optimal temperature when operating with high slag
rates. The increased coke use elevates a mill’s emissions, which can result in penalties and
possible shut-downs.

Metallurgical coal prices

Beyond the threat of punishment for exceeding emissions thresholds, mills also have an
obvious financial motivation to minimize coke consumption, particularly when prices
are high. With other factors in balance, we would expect therefore to see a positive
correlation between the price of coking coal (the pre-cursor to coke) and the 65-62% Fe
spread (Figure 3).

Fig. 3: CFR China Premium Hard Coking Coal Price vs 65-62% Spread

Higher coke prices incentivize increased usage of higher quality ores, which decrease blast
furnace slag rates. The higher metallurgical coal prices of recent years have been another
factor contributing to wider quality spreads in iron ore.

350 35
300 30

250 25
20
200
15
150
10
100
5
50 0
0 -5
Oct-16

Oct-18
Oct-15
Oct-14

Oct-17
Jan-16

Jan-18
Jan-14

Apr-16

Apr-18
Jan-15
Apr-14

Apr-15

Jan-17

Apr-17
Jul-16

Jul-18
Jul-14

Jul-15

Jul-17

Fastmarkets MB PHCC ($/tonne, CFR China) [LHS]

Iron Ore 65-62% Fe Index Differential ($/tonne) Source: Fastmarkets MB

Steel quality and production efficiency

Another policy-led factor that has contributed to the favorability of higher grade ores has
been the government’s push to focus more on the quality of steel the country produces as
opposed to just the quantity.

This effort has seen the mass shut-down of small-scale, scrap-fed induction furnaces that
were illegally producing sub-standard steel (mainly rebar). It goes further though with
‘capacity swap’ quotas encouraging mills to replace smaller furnaces with bigger, more
efficient ones – the share of ‘large’ (>4,000 cubic meters) blast furnaces having increased
from 15% in 2013 to over 22% in 2018. Larger blast furnaces are more energy-efficient and
tend to produce higher and more consistent-quality steel.

4
This trend has an impact on demand for raw materials though. Larger furnaces require
both higher quality iron ore and higher quality coke, with the strength to support a
heavier burden and maintain permeability for efficient gas flow. In the debate over
whether the flight to higher grades represents a cyclical or structural change, the
construction of bigger furnaces undoubtedly supports the latter case.

Supply ratio of different ore grades

While the lion’s share of inter-grade spread expansion since the start of 2016 can be
attributed to China’s industrial reforms and strong steel demand, the ratio of ore
production across the grade spectrum is also an important variable in determining the
relative value of different products.

A change in the production ratio across the grade spectrum, driven for example by
operational outages, suspensions, new starts or ramp-ups, affects the chemistry of the
‘global product mix’ – a virtual blend of all the world’s ores. A change in the chemistry of
the global product mix often triggers changes in purchasing behavior and the penalties
applied by the market to certain impurities.

A tilt in the silica to alumina ratio in the global product mix can determine which
impurity consumers are more sensitive towards by affecting the degree of flux addition
required to maintain a consistent basicity ratio (CaO:SiO2) in the blast furnace.

Too silica rich, and extra silica will require additional CaO flux to balance the basicity
ratio, which further increases slag rate. Too alumina-rich though and any extra alumina
needs to be diluted through addition of silica, which in turn requires more flux to balance
the basicity ratio. This more compound slag rate increase in an alumina-constrained
system therefore tends to result in a more rapid escalation of penalties for alumina.

The recent profit-driven productivity focus has amplified the price adjustments for
impurities, with record levels being seen for silica in 2017 and alumina in late 2018 (Figure
5). This swing in sensitivity has been influenced in large part by fluctuations in Chinese
domestic concentrate production. China’s lack of naturally enriched iron ores means
that it mainly produces high-silica concentrates.

A price-incentivized resurgence of output of this material in 2017 drove the temporary

surge in silica penalties, while a dramatic drop due to stricter environmental measures in
2018 has contributed to the record alumina penalties we have seen recently.

5
Fig. 4: China’s Domestic Iron Ore Output

While China’s steel industry reforms have prompted an increased aversion to impurities
in general, it’s been the variability in high-silica domestic concentrate supply that has
primarily driven the fluctuation between silica and alumina sensitivity specifically.
Source: National Bureau of Statistics of China, Fastmarkets MB

2000 6
5
1500
4
1000 3
2
500
1
0 0
2004 2006 2008 2010 2012 2014 2016 2018e

ROM Production (Mtpa) Concentrate (66% Fe basis)

ROM to Concentrate Ratio [RHS]

Fig. 5: Evolution of Grade Spreads and Impurity Penalties

120 0
-1
100
-2
80 -3
-4
60
-5
40 -6
-7
20
-8
0 -9
Aug-15 Feb-16 Aug-16 Feb-17 Aug-17 Feb-18 Aug-18

MBIOI-62 [LHS] MBIO-58 [LHS] MBIOI-65 [LHS]

Si-VIU ($/1%) Al-VIU ($/1%) P-VIU ($/0.01%)
Source: Fastmarkets MB

The way changes in impurities impact the inter-grade spreads, and particularly the
65-62% Fe differential can be complex. Any impurity-sensitive conditions will support
a wider 65-62% Fe spread given the specification differences between the two indices.
However, the tendency for alumina penalties to escalate more radically than those for
silica mean the 65% Fe specification may typically hold a larger premium to the 62%
under the most alumina-sensitive conditions.

6
Development of a high grade
derivative contract

A natural product of variable quality, iron ore has always been a challenging market
to commoditize. Through a concerted and coordinated effort nearly a decade ago the
industry managed to sufficiently coalesce pricing around the 62% Fe benchmark grade
to allow an important step in the market’s maturation – the development of a liquid
exchange-based derivative contract.

But 2009 (the year SGX launched its first iron ore swaps contract) was arguably a simpler
time with smaller, more stable value relationships between different grades of ore.

Greater market complexity, combined with more sophisticated approaches to value

determination, has seen increased diversification in pricing mechanisms, diminishing the
status the 62% Fe grade once held as ‘the’ iron ore price.

Fig. 6: Evolution of Grade Spreads and Impurity Penalties

Make iron ore great again? Google trend analysis for worldwide searches with ‘iron ore’
as a key term shows a dwindling level of interest in the sector in recent years, perhaps in
response to lower volatility of the benchmark 62% Fe index and its decreased capacity
to reflect prices across the whole iron ore product spectrum. Meanwhile though the
physical market has experienced one of its most intriguing periods, with more dynamic
differentials emerging between different grades and products. The development of
a high-grade derivative may prove to be the catalyst that re-captures the interest of
the financial mainstream by better reflecting the nuances of a market that has been
anything but boring.

160

140

120

100

80

60

40

20

0
Aug-13 Feb-14 Aug-14 Feb-15 Aug-15 Feb-16 Aug-16 Feb-17 Aug-17 Feb-18 Aug-18

MB65 MB62 65-62 Spread

7
The first efforts to introduce a paper market for another iron ore benchmark grade
came in 2015 with the launch of 58% Fe contracts on the SGX and CME. However, the
timing of the development in the midst of a painful price downturn found many market
participants in survival mode, lacking the appetite to drive liquidity growth while focus
was locked on cost-reduction and operational efficiency.

The fate of the low-grade derivatives may in part have spooked exchanges from making
forays into other grade segments in subsequent years. However, market participants
point to a number of differences that may bode favorably for the high-grade contract.
First and foremost, as the settlement benchmark for Vale’s IOCJ, there is more volume
directly tied to the Fastmarkets MB 65% Fines Index than there ever was to any low-
grade index.

At approximately 130Mtpa, IOCJ is one of the world’s largest single product streams by
volume, and certainly the market’s largest by value, meaning physical players around
the world are routinely exposed to huge financial risk with no access to a viable hedging
mechanism.

The industry also points to the scope for a 65% Fe derivative to serve as a hedging option
for the wider high-grade market segment as something that will broaden its appeal.
Estimates put the volume of global iron ore exports grading at 63.5% Fe or above at
around 500Mtpa, and the percentage of the total seaborne market exceeding 65% Fe at
around 37-38%. This volume comprises high grade fines, lumps, concentrates and pellets
– all of which have distinctive pricing dynamics. However, a 65% Fe fines derivative would
constitute a significantly more value-aligned tool than is currently offered by the 62% Fe
paper contract.

Assessing the viability of the 65% Fe derivative as a hedging tool for other products
in the high grade sector, the ~99% correlation (Figure 7) shown with the 66% Fe
Concentrate Index (MBIOI-CO) points towards the new contract being a very credible
option for managing exposure to concentrate and pellet feed prices. The stronger
correlation shown with the pellet price (MBIOI-PT) also backs up the logic of the current
market negotiations to shift 2019 pellet contracts away from the 62% Fe price to use the
65 as the base price.

8
Predicting future movements of
inter-grade price differentials

Carrots and sticks

Given the number of moving parts that combine to determine inter-grade price spreads,
anticipating their future expansion or contraction won’t be straightforward. Some
market participants hold the view that the long expansion in the 65-62% Fe differential
is due to give way to a prolonged contraction, but an analysis of the drivers suggest we
could be entering choppier waters.

Post-2016 the factors driving the inter-grade spreads have for the most part all been
pulling in the same expansionary direction; stronger mill margins due to capacity
cuts and robust domestic steel demand, stricter environmental regulation, higher
metallurgical coal prices, a focus on steel quality over quantity, and a less balanced
impurity profile in the global product mix.

In other words there’s still something of a carrot and stick debate, since so far the
economic incentives have motivated mills to operate in a fashion that’s fairly well
aligned to the stricter environmental policies. The uncertainty will come when these
spread-driving factors start to move in opposition, as they have begun to.

Fig. 7: One of the strongest relationships between any two Fastmarkets iron ore benchmarks is the ~99% correlation
between the MBIOI-65-BZ and the MBIOI-CO – the 66% Fe Concentrate Index. Another high-grade product segment,
pellet, also correlates more strongly with the 65% Fe index than with the 62%, suggesting the new SGX contract could
well constitute a more viable hedging option than the 62% Fe contract for the high-grade market segment as a whole.
Source: Fastmarkets MB

Pellet
(BF & DR)

High
High Grade Conc &
Grade
Iron Ore Pellet Feed
Fines

High
Grade
Lump

9
In recent years we have witnessed bumpy periods in the 65-62% Fe differential in
response to fluctuations in mill margins, coke prices and impurity sensitivities. But a
serious profitability plunge under today’s more draconian environmental restrictions
is, for example, a scenario yet to be truly tested. Certainly, the stricter anti-pollution
legislation and shift in steel quality requirements are structural changes that should
support a higher floor for the 65-62% Fe differential, but what that base value may be
under less financially favorable market conditions is a major unknown.

Demand

Perhaps the most significant factor that could drive a change in the 65-62% Fe
spread may be a continued drop in steel demand in China, which would negatively
impact mill margins. Arguably just as important a factor as the capacity cuts and
environmental policies, the strength of end-user steel demand in recent years is not to
be underestimated. The drop in China’s steel exports from their peak at over 110Mt in
2015 to just 75Mt two years later in 2017, and the domestic market’s ability to absorb
record crude steel production volumes at such high rates of profitability, is testament to
the recent health in the country’s industrial economy.

However, good times don’t last forever and some analysts believe that Chinese steel
demand may have peaked. For example, Fastmarkets MB Research predict a 1.6% drop
in crude steel output in 2019 relative to 2018, with the shrinkage extending further into
the future for a compound annual growth rate of -0.4% out to 2030. Practically though,
the range of future scenarios could still be rather broad.

Taking the more bearish views at face value, a protracted waning of demand and
subsequent drop in steel prices would deal a heavy blow to mills’ profit margins and put
contractionary pressure on the iron ore quality spreads.

Fig. 8: Chinese Crude Steel Output (Quarterly forecast)

Chinese crude steel expansion has been strong since Q2’16 following economic stimulus-
led rally. Inventory-driven correction in crude output expected through 2019. Source:
Published in Fastmarkets’ Iron Ore Market Intelligence Service

250 Crude steel output Y-o-y change [RHS] 10%

240 8%
230
6%
220
4%
Million tonnes

210
200 2%
190
0%
180
-2%
170
160 -4%

150 -6%
Q3'16
Q4'16
Q1'17
Q2'17
Q3'17
Q4'17

Q3'18
Q4'18

Q3'19
Q4'19

Q3'20
Q4'20
Q1'21
Q2'21
Q3'21
Q4'21
Q1'22
Q2'22
Q3'22
Q4'22
Q2'20
Q1'16

Q1'19
Q1'18

Q1'20
Q2'16

Q2'19
Q2'18

10
Indeed we have already seen a sharp fall in the 65-62% Fe differential, which has almost
halved between the start of September and the start of December in response to a drop
in HRC prices, closely followed by similar slides in rebar.

The recent sell-off in steel looks to have roots in the financial markets, and speculations
of weakening demand thanks to the US/China trade war in particular. Fundamentals
meanwhile, at least in the longs segment, still look relatively robust with export and
inventory volumes both low, new floor space starts close to recent highs, and buoyant
property prices proving there is still air in the real-estate bubble. The only cause for
some concern might be the pace of growth, with infrastructure investment growth – a
potential leading indicator of longs demand – slowing this year posting a 3.7% year-on-
year increase in the January-October period of 2018, compared with a 19.6% increase
at the same point in 2017. Flats are still a different story though, having been under
pressure for some time due to a decline in automotive sector output, and weaker
international demand for white goods exports as a result of the trade war.

The varying fortunes of producers across the steel product spectrum should serve as a
reminder that ‘mill margins’ as a single factor is an oversimplification of a very complex
gauge. In practice, a more granular approach to evaluating profitability across the steel
industry is necessary to interpret the implications for raw material purchasing behavior.

Supply

On the supply side of the equation, a lackluster output response to price premiums in
the high grade sector up to now has arguably kept prices higher for longer. However,
bringing new high grade supply on line is neither quick nor simple. Though iron is the
fourth most abundant element in the Earth’s crust, naturally enriched and economically
accessible high grade deposits are a rare commodity. Outside of Vale’s S11D ramp up
(part of which will likely go into the company’s mid-grade Brazilian Blend) and the
re-start of Mount Gibson’s Koolan Island mine there is little else in the way of high
grade DSO expected to enter the market in the medium-term. While Simandou’s
potential 100Mtpa of 65% Fe fines would hypothetically be a genuine game changer, its
development looks as far off as ever.

A line-up of beneficiated ore projects (both brownfield and greenfield) looking to target
the pellet feed and sinter concentrates sector look like strong candidates on paper to
hit the market in the next few years, but the high-CAPEX and long lead-time tendency
of these projects has to be taken into account. Anglo American’s Minas Rio project,
which has been suspended since March 2018, has recently resumed production in late
December. A restart at Samarco probably won’t be seen before 2020 though, and even
then is unlikely to reach previous run-rates.

Perhaps the most impactful supply response in the short term may come from efforts
to incrementally improve specifications of existing product streams. The successful
reduction of silica levels in certain products in late 2017 was a contributing factor in
declining silica penalties, and efforts to improve specifications with respect to alumina
could similarly soften penalties and promote a contraction in the 65-62% Fe spread.

11
Geological constraints though, as well as the likely continued suppression of Chinese
domestic concentrate supply, may affect the degree to which the alumina to silica
ratio of the global product mix can be meaningfully lowered in coming years. In fact,
with mines such as Yandi coming to the end of their operating lives, many believe that
greater alumina and phosphorus sensitivity could be a longer-term feature.

Environmental policy

The strong hand of the Chinese government is once again being felt by the market
recently with the announcement of this winter’s restriction policies on steel production.
After blanket curbs last winter the policies this year appear more pragmatic, with a
case-by-case approach focused on pollution reduction permitting more flexibility around
the means that achieve the ends.

The lighter touch approach has already had an impact on ore purchasing behavior, lower
grade products finding some favor once more – especially CID ores that represent a
cheaper option to suppress alumina levels in a less productivity-concerned environment.

Don’t mistake these shorter-term strategy tweaks as a weakening of China’s anti-

pollution efforts though. We have already seen emergency production curbs ready to be
implemented in areas like Tangshan, when emissions targets look under any threat.

Scrap

Another factor expected to shake up the industry is China’s growing use of ferrous scrap
as a steel feedstock. This potential ‘scrapocalypse’ for iron ore demand underpins many
of the more bearish long-term views for the industry, but the potential timing and extent
of an increase in scrap substitution is anything but clear.

Profitability and environmental incentives in recent years have prompted mills to use
more scrap, despite its high price relative to the cost of producing pig iron (Figure 9).
Indeed, ratios of crude steel to blast furnace iron production indicate that a large
proportion of the recent steel output increase can be attributed to higher rates of scrap
consumption.

The question is whether a high BOF scrap rate will persist in a scenario of weakening
steel demand and dented profit margins, given its uncompetitive cost. Despite the
environmental push, Chinese mills have historically shown themselves to be cost-
sensitive and, unless prices drop, would likely look to reduce scrap rates in the BOF.

12
Fig. 9: Scrap has priced at a premium to hot metal production cost in China
during 2018

The price spread between ferrous scrap and a hot metal proxy in China has been positive
so far in 2018. Source: Fastmarkets MB Research

450 200

400
150
350

300 100

250
$/t

$/t
50
200

150 0

100
-50
50

0 -100
Jan-14 Oct-14 Jul-15 Apr-16 Jan-17 Oct-17 Jul-18

Hot metal over scrap premium, $/t [RHS]

Hot metal proxy

China domestic scrap, $/t del, excl. VAT

Beijing’s support for capacity swaps to EAF-route production may in the short term
delay a drop in scrap prices to converge with hot metal production costs. But with
China’s obsolete supply predicted to surge from 60Mt in 2017 to reach anywhere
between 168Mtpa (Worldsteel projections) and 340Mtpa (McKinsey estimates) by 2030,
the longer-term scrap price outlook may potentially look more competitive – especially to
anyone skeptical of the rate of EAF adoption in the country.

Worldsteel forecasts that China’s share of BOF-route production will continue to drop
from 92.7% in 2017 to 89% by 2023, with Fastmarkets MB Research predicting a further
fall to 82% by 2030 – a retraction yes, but not a rapid and industry-wide transition
by any means. Many point to China’s limited (at least for now) availability of cheap,
clean energy, needed for EAFs to be both cost-competitive and more environmentally
friendly than BF/BOF route production, as a counter-argument to a major shift. If EAF
penetration does fail to meet more bullish expectations then scrap may become cheap
enough to compete with pig iron for BOF share even if crude steel production falls – a
threatening scenario for iron ore demand.

13
Fig. 10: Several scenarios could play out for China’s future scrap availability and consumption rate, each with different
implications for the iron ore market. This table summarizes the potential impact on iron ore’s inter-grade differentials in
four scenarios, varying with respect to mill profitability and scrap price competitiveness. It’s worth noting that if overall
iron ore demand decreases we will likely see cost-curve pressure return, the result of which may influence inter-grade
spreads from the supply side of the equation.

STRONG  MILL  MARGINS

Expensive  scrap Cheaper  scrap
Crude  Steel  Output Scrap  Consumption Iron  Ore  Demand Inter-­‐Grade  Differential Crude  Steel  Output Scrap  Consumption Iron  Ore  Demand Inter-­‐Grade  Differential

Status  quo:  steel  demand  stays  strong,  mills  profitable,  and  crude  steel  output  rates  high.  Mills  continue  to  use  high   Strong  steel  demand  and  profit  margins  maintain  incentivize  for  high  crude  steel  output.  Cheaper  scrap  due  to  
scrap  rates  to  support  productivity  in  spite  of  high  prices.  Iron  ore  demand  remains  stable  and  high,  with  focus  on   expectation-­‐exceeding  supply  surge  encourages  capacity  switches  to  EAFs.  Overall  iron  ore  demand  drops,  but  
quality  supporting  wide  inter-­‐grade  spread. quality  pickiness  perists  to  maximize  yield  from  remaining  BF  capacity.
WEAK  MILL  MARGINS
Expensive  scrap Cheaper  scrap
Crude  Steel  Output Scrap  Consumption Iron  Ore  Demand Inter-­‐Grade  Differential Crude  Steel  Output Scrap  Consumption Iron  Ore  Demand Inter-­‐Grade  Differential

Drop  in  steel  demand  drags  down  mill  profit  margins,  incentivizing  a  reduction  in  crude  steel  output.  Cost-­‐sensitive   Crude  steel  output  drops  in  response  to  demand  &  margins  weakening.  Supply  surge  cheapens  scrap,  improving  cost-­‐
mills  reduce  use  of  expensive  scrap  feed,  using  higher  percentage  of  pig  iron  in  BOF.  Reliance  on  BF  iron  should   competitiveness  vs  pig  iron.  Mills  maintain  scrap  consumption  with  iron  ore  bearing  the  brunt.  Lower  yield  rates  
soften  demand  blow  to  iron  ore  &  support  quality  spreads. required  from  BFs  favour  lower  grade  ores,  narrowing  inter-­‐grade  spread.    

The verdict

Weighing the drivers, the near-term outlook does appear generally indicative of a 65-
62% Fe spread contraction from recent record highs. However, a sharp fall in recent
months may have seen much of the damage already done, and could even have laid the
ground for a potential rebound. In truth, the real lesson may be that with cyclical and
structural spread-driving factors now clashing with one another, the outlook is likely to
be choppier and less predictable.

Index Analytics

Inter-grade arbitrage

Despite the fragmentation of the market, the link between various iron ore grades and
products is more pronounced than it might appear at first glance. The benchmark
indices for high, mid and low grade ores show an interactive relationship connected
to the physical arbitrage opportunities of blending, and for the most part, high-grade
premiums and low-grade discounts mirror each other’s movements. An arithmetic blend
of the 65% and 58% Fe indices weighted to proxy a 62% Fe product (a simplification
not accounting for gangue chemistry) tracks in close alignment with the 62% Fe index
itself (Figure 11), revealing the efficiency of valuation across the grade spectrum, with
deviations from long term mean indicating a potentially exploitable blending arbitrage.

This feature of the market is more than just a curiosity; it may play into projections
of the future 65-62% Fe differential. It’s widely understood, for example, that a
disproportionate volume of stockpile volume at Chinese ports is of lower grade given the
comparative favourability of higher grades in recent years. This could in theory buffer an
improvement in low-grade prices even in conditions favouring a narrowing of discounts,
thus helping maintain a wider high-grade premium for longer than expected.

14
Fig. 11: Physical arbitrage indicator - 62% Fe vs 65/58% Fe blend

High-grade premiums and low-grade discounts tend to mirror each other’s movements, with a weighted average of
the 65% and 58% Fe indices, approximated to proxy a 62% Fe blend, tracking very closely with the 62% Fe index itself.
Deviations from the mean tend to revert with time as they represent something of a value-in-use arbitrage opportunity.

160 8
140 6
120 4
100
2
80
0
60
40 -2

20 -4
0 -6

8
14

14

15

15

16

16

17

17

18

18
1

-1
1

-1
1

-1

-1
1

-1

n-

r-

l-
n-

r-

l-
n-

r-

l-

n-

r-

l-
n-

r-

l-

ct
ct
ct

ct
ct

Ju
Ju
Ju

Ju
Ju

Ap
Ap
Ap

Ap
Ap

Ja
Ja
Ja

Ja
Ja

O
O
O

O
O
MBIOI-62 65/58% Blend MBIOI-62 vs 65/58 Blend Mean (MBIOI-6262 vs 65/58 Blend)

Source: Fastmarkets MB

Technical analysis & correlation with sector variables

In seeking to better understand the behavior of any index, traders will naturally look for
correlations with related markets and sector variables – the holy grail being the discovery
of leading indicators of price movement. In this section we analyze the relationships
between movements in iron ore prices and inter-grade differentials, and trends in
macroeconomic indicators.

The MBIOI-65-BZ Statistics

60
50
40
30
20
10
0

-10 -5 0 5 10
Percentage change

Visually, the graph showing the continuous probability distribution (kernel density
estimation) of MBIOI-65-BZ daily returns suggests similarity in the distribution
of positive and negative percentage changes over time. However, the Shapiro-
Wilk test shows that the index behavior does not follow a normal distribution. A
positive skewness coefficient of 1.21 means that the right tail of the distribution
is longer relative to the left tail – the result of a few highly positive percentage
moves. Indeed, a kurtosis coefficient of 12.89 for the index (much larger than normal
distribution coefficient of 3) means that MBIOI-65-BZ has tended to experience a
higher frequency of occasional extreme returns than in markets that show normal
distributions in price movement. This is of course somewhat expected looking at the
iron ore market – the roller coaster price rides in which are certainly not for the faint
hearted, and which make risk management all the more important.

15
Economic data like GDP, inflation metrics such as CPI and PPI, money supply and interest
rates, and sentiment indices like manufacturing PMI, are all indicators that could
factor into iron ore price trends. Production indicators like steel output and industrial
production, and property sector indicators like house prices, sales and investment, may
also show correlations.

As can be seen from the correlation matrix below (Figure 12), the indicators that display
the highest correlation with iron ore prices (both 62% and 65% Fe) are accumulated
growth rates in real estate investment, industrial production and steel production.
Interestingly, the growth rate of steel production shows a significantly stronger
correlation with the price of 65% Fe iron ore than it does with 62% Fe.

This relationship can be easily explained by the drivers discussed earlier in this white
paper – the higher efficiency and increased rate of steel production being a key factor
behind the relative recent strength of high-grade ore prices.

Cross-correlation analysis can help show us whether these highest-correlation indicators

either lead or lag the movements of the MBIOI-65-BZ. In the case of real estate investment
growth rate, lag 0 has the highest correlation with the 65% Fe index value, implying
approximately simultaneous change. Similarly, changes in the growth rate of industrial
production and in the 65% Fe index value also appear simultaneous, with lag 0 again
showing the highest correlation.

Fig. 12: Cross-correlation analysis indicates that the movement of the MBIOI-65-BZ
shows a 1-month lag in correlation to crude steel production, while the correlation
is approximately simultaneous with respect to industrial production and real estate
investment. Source: Fastmarkets MB, various (see below)

16
The MBIOI-65 & Real estate investment %

0.8
0.6
0.4
ACF
0.2
0.0
-0.2

-15 -10 -5 0 5 10 15
Log

The MBIOI-65 & Steel Production

0.6
0.4
ACF
0.2
0.0
0.2
-0.4

-15 -10 -5 0 5 10 15
Log

The MBIOI-65 & Industrial Production

0.8
0.6
0.4
ACF
0.2 0.0
-0.2

-15 -10 -5 0 5 10 15
Log

17
However, in the case of steel production rate growth, the highest correlation was
reached at lag 1. That suggests that that steel production growth statistics have tended
to moves one month ahead of a correlated reaction in the 65% Fe index price, and may
imply a potential leading indicator.

If we look specifically at the correlations between economic indicators and the 65-62%
Fe price spread (Figure 13) we find some of the strongest relationships again with crude
steel production, but also this time with manufacturing PMI, CPI and money supply.

Again, cross-correlation analysis shows that the correlations are strongest at different
time lags. Steel production growth rate statistics once again correlate most strongly
with the 65-62 spread at lag 1, meaning the spread reacts with an approximately one
month delay. However, both PMI and money supply (M2) correlate most strongly at lag
2, suggesting they are potentially earlier leading indicators of movement in the 65-62
index spread.

Fig. 13: Cross-correlation analysis indicates that the movement of the spread between
the 65 and 62% Fe indices shows a 1-month lag in correlation to crude steel product,
while an approximately 2-month lag is shown in its correlation with manufacturing PMI
and money supply data. Source: Fastmarkets MB, various (see below)

18
Spread & Steel Production

0.6
0.4
ACF
0.2 0.0
-0.2

-15 -10 -5 0 5 10 15
Log

Spread & M2
0.8
0.6
0.4
ACF
0.2 0.0
-0.2

-15 -10 -5 0 5 10 15
Log

Spread & PMI

0.6
0.4
ACF
0.2 0.0
-0.2

-15 -10 -5 0 5 10 15
Log

19
 Of course, correlation does not necessarily imply causality. But interestingly, one possible
explanation for the results builds a picture of the potential cause and effect dynamics
of the market. For example, an announcement of money supply data may trigger a
change in purchasing manager sentiment, which then translates into a change in steel
production rate, before being felt in the seaborne raw materials markets as quality
differentials react in response to changing demand. The reactivity of iron ore prices in
response to steelmaking costs and incentives, as well as to macroeconomic data and
news triggers, creates an interesting market where information and understanding
can be an edge in maximizing value. The development of a high-grade derivative to
complement the existing 62% Fe paper market is a long awaited innovation, which will
allow the market to apply more multi-dimensional trading strategies that closer reflect
the complexities of the diverse physical iron ore market.

Correlation matrix data:

• MBIOI 65 – MB Daily Iron Ore Index, 65% Fe, CFR China, $/dmt
• MBIOI 62 – MB Daily Iron Ore Index, 62% Fe, CFR China, $/dmt
• % Real estate - Investment of Real Estate, Accumulated Growth Rate (%) (National Bureau of Statistics of China)
• % IP – Chinese Industrial Production, Accumulated Growth Rate (%) (National Bureau of Statistics of China)
• % SP – Chinese Crude Steel Production, Accumulated Growth Rate (%) (National Bureau of Statistics of China)
• PMI – Chinese Manufacturing Purchasing Managers’ Index (National Bureau of Statistics of China)
• % M2 – Chinese M2 Supply year-on-year growth (National Bureau of Statistics of China)
• PPI – Chinese Producers Price Index (The same month last year=100) (National Bureau of Statistics of China)
• CPI – Chinese Consumers Price Index (The same month last year=100) (National Bureau of Statistics of China)
• % Housing sales – Chinese Commercialized Buildings Sales, Accumulated Growth Rate (%) (National Bureau of Statistics of China)

20
Iron ore
What we do Prices
We understand the importance of iron ore to the global raw materials and
steel markets. We provide a range of price assessments and indices, including
industry benchmarks, from around the globe covering grades 58%, 62%, 65%
and 66% Fe.

Empower yourself with the most market reflective prices, historical trends and
future outlooks with our unparalleled coverage.

Our price assessment process

Being transparent about our price assessment and index process is important
to us. You can be confident that our pricing process is impartial and market
reflective.

Our methodology is aligned with core IOSCO principles and we have

successfully completed an assurance review (from PricewaterhouseCoopers)
for our financial benchmarks.

1 2
Trained price reporters Market data is collected
in 13 global locations provide specialist from numerous approved contributors
market knowledge from across the supply chain

4 3
Data analysis Pricing data is entered into our
Data is analyzed and, if required, specially designed pricing database,
normalized. Outliers are removed. MInD (Markets Information Database),
to ensure a full audit trail

5 6
Our integrated peer-review system Data is published through Price Book
ensures that all data is verified by at least and our data solutions, which are made
two reporters & an editor or senior journalist available to subscribers who can be
to ensure quality control & compliance confident about the quality of our prices

21
Market news
Our market news offers global coverage across all steel making and raw
material markets, with a dedicated expert team for each.

From breaking news and market summaries to in-depth industry analysis

and expert commentary, we guarantee complete coverage of your
market globally.

Forecasting and analysis

Our independent research helps businesses worldwide to minimize risk
and anticipate future trends, based on insights and analysis from experts
who specialize in their markets.

Our forecasting & analysis services include:

• Steel Raw Materials Market Tracker
• Iron Ore Market Intelligence Service

Learn more about our iron ore coverage

Get in touch today to EMEA: +44 (0)20 7779 8260

Asia: +65 62 382 139
find out how we can Americas: +1 312 366 2541
help your business [email protected]
www.fastmarkets.com