Content      ..     1      2      3      4      ..

INTERNATIONAL MONETARY FUND. Navigating Global Divergences (October 2023) - page 3

WORLD ECONOMIC OUTLOOK: Navigating Global Divergences

Box 2.2. Fiscal Imprudence and Inflation Expectations: The Role of Monetary Policy Frameworks

Fiscal imprudence—high levels of public debt to GDP—

Figure 2.2.1. Inflation Expectations in

is generally regarded as having the potential to generate

Emerging Market and Developing Economies:

uncertainty and influence inflation expectations by

Monetary Policy Frameworks and Public

eroding perceptions of monetary policy credibility and

Debt Interactions

independence.¹ That much has been clear since the work

(Percent)

of Sargent and Wallace (1981) and Leeper (1991).

This box empirically examines how the level of inflation

Lower debt to GDP

Higher debt to GDP

expectations is related to an economy’s monetary policy

Higher debt to GDP plus persistent deficit

framework, given the level of public debt.

Distribution in 2007 (right scale)

Distribution in 2021 (right scale)

In the study presented in this box, the soundness

10

1. Government Debt

0.7

of monetary policy frameworks is captured by a

0.6

new index, the IAPOC index, developed by Unsal,

8

0.5

Papageorgiou, and Garbers (2022).² It shows that even

after economy-specific controls and time-invariant

6

0.4

characteristics are accounted for, higher public debt

0.3

4

is associated with expectations of higher inflation,

0.2

given a specific level of monetary policy framework

2

(Figure 2.2.1, panel 1).³ Tis heightened impact is

0.1

even more evident when the focus is on the stock of

0

0.0

public debt in foreign currency and exacerbated when

0.0

0.2

0.4

0.6

0.8

1.0

fiscal deficits are persistent (Figure 2.2.1, panel 2).

20

2. Government Debt in Foreign

0.7

However, as monetary policy frameworks are improved

Currency

0.6

(as seen with the shift in the IAPOC index distribu-

15

tion in emerging market and developing economies

0.5

over the past 15 years), inflation expectations become

0.4

10

less sensitive to the level and composition of public

0.3

debt or persistent fiscal deficits.

0.2

Overall, the study findings indicate that difficulties

5

posed by higher public debt levels for managing infla-

0.1

tion expectations in emerging market and developing

0

0

economies could be eased by adopting strong monetary

0.0

0.2

0.4

0.6

0.8

1.0

policy frameworks. Whereas monetary policymaking

Sources: Consensus Economics; Unsal, Papageorgiou, and

Garbers (2022); and IMF staff calculations.

Te authors of this box are Omer Akbal, Mariarosaria Comu-

Note: Horizontal axes show the IAPOC index level. The lines

nale, Marina Conesa, Chris Papageorgiou, and Filiz Unsal.

show the marginal effects of monetary policy framework

¹See Brandao-Marques and others (2023) for a recent empir-

changes (according to the IAPOC [Overall Monetary Policy

Framework] index) on mean inflation expectations,

ical study of the issue and Bianchi and Melosi (2022), Bianchi,

conditional on the ratio of total (foreign-currency)

Faccini, and Melosi (2022), and Cochrane (2022) for theoreti-

government debt to GDP. “Higher (lower)” debt is the

cal arguments.

average debt to GDP, conditional on its being above (below)

²Te IAPOC index and its subindicators quantify the

the sample mean. Estimates are from a fixed-effects panel

soundness of monetary policy frameworks across countries

regression across economies of mean inflation expectations

through three pillars: Independence and Accountability

on the interaction of the IAPOC index score and debt to GDP.

(I and A), Policy and Operational Strategy (P and O), and

Distributions represent the density of the IAPOC index for the

Communications (C). Tis comprehensive index enables

assessed economies in 2007 (dashed) and 2021 (solid), with

a multidimensional characterization of monetary policy

a rightward shift indicating improvement.

frameworks—going beyond monetary policy or exchange rate

regime classifications—across 13 advanced economies and

in many of these economies is better equipped than

37 emerging market and developing economies. See Unsal,

15 years ago to serve as an anchor of stability, the adop-

Papageorgiou, and Garbers (2022) for further details. Te

tion of a prudent fiscal policy approach remains key to

data set has been updated to 2021.

effective preparation for challenges and to prevent the

³Advanced economies do not show this differential sensitivity

to debt levels over different IAPOC index scores.

risk of fiscal dominance in the future.

66

International Monetary Fund | October 2023

CHAPTER 2 Managing Expectations: Inflation and Monetary Policy

Box 2.3. Energy Subsidies, Inflation, and Expectations: Unpacking Euro Area Measures

Several European economies have used energy subsidies,

Figure 2.3.1. Marginal Impacts of Fiscal

tax cuts, and price caps to help smooth the impact of

Measures for Relief from the Energy Price

recent shocks to energy prices on incomes and inflation.

Shock on Inflation and Expectations

The effectiveness and desirability of such measures depends

(Percentage point deviation from no-measures

on many factors beyond the scope of this box, including

scenario)

their impact on energy markets, resource misallocation,

3

1. Channels

and fiscal sustainability, as well as details of the policy

Headline

design. One important channel is inflation expectations.

Direct

Temporary energy subsidies directly lower inflation today

2

Indirect effects

One-year-ahead core inflation

but increase it relative to the no-measures scenario after

they expire, smoothing the overall inflation path. If energy

1

subsidies are perceived as temporary, the expectations

channel may reduce their effectiveness in lowering infla-

0

tion, as expectations of higher future inflation may affect

price-setting today.

-1

2022

23

24

25

A model from the IMF’s Flexible System of Global

Models is used in this box to simulate the impacts on

2.0

2. Headline

3. Core Inflation

2.0

Expectations

expected and realized inflation of announced energy

1.5

1.5

relief measures (price subsidies and caps) in the euro

1.0

1.0

area.¹ Te simulation assumes that the sharp upward

0.5

0.5

shock to energy prices in 2022 is temporary and

0.0

0.0

unwinds. It also includes the indirect effects of energy

-0.5

-0.5

prices on core inflation through the supply chain. Te

-1.0

By perceived relief persistence:

-1.0

model estimates that fiscal relief measures lowered

Baseline, as announced

-1.5

-1.5

euro area inflation by 0.9 percentage point in 2022

Alternative, longer by one year

and by half a percentage point in 2023 (Figure 2.3.1,

-2.0

-2.0

2022

23

24

25

2022

23

24

25

panel 1). Although additional fiscal borrowing to

finance subsidies boosts demand, its effect on core

Sources: Dao and others (2023); and IMF staff calculations.

inflation is more than offset by the reduction in

Note: Panel 1 shows the marginal impacts on inflation of

supply-chain costs. Tese fiscal measures smooth out

announced fiscal relief measures for energy, using the IMF’s

Flexible System of Global Models. The blue bars show the

the inflation impact of the energy shock over time,

direct effects of measures (subsidies, tax cuts, or price caps

leading to a rise in inflation over 2024-25 (relative to

on consumer energy prices), and the red bars show the

the no-measures scenario) and preventing an under-

indirect effects from changes in aggregate demand, supply

chain costs, and core inflation expectations. The baseline in

shoot as energy subsidies expire and the energy shock

panel 2 assumes fiscal relief measures last in 2022 as

unwinds. Te measures have a net neutral effect on

originally announced. The alternative assumes that

core inflation expectations in 2022 but increase them

households misperceive and expect measures will last

longer, but then in 2023 they realize their error and adjust to

by 0.7 percentage point over 2023-24. Tese find-

the announced path.

ings assume, however, that agents fully understand

the temporary nature of the subsidies. What if agents

misperceive and think that the subsidies will last

prices by more in 2022, because they now expect core

for a year more than announced? In this alternative

inflation to be lower in 2023. Te fall in inflation

scenario, expectations fall more in 2022 (Figure 2.3.1,

expectations increases the impact of fiscal policy on

panel 2). Subsidies also lead firms to lower their

inflation from -0.9 to -1.1 percentage points in 2022

and from -0.5 to -0.6 percentage point in 2023. But

once agents realize their error and correct, inflation

Te author of this box is Chris Jackson.

¹See Dao and others (2023) for further details on the structure

and expectations bounce back, highlighting the role of

of the model and simulation calibration.

the expectations channel.

International Monetary Fund | October 2023

67

WORLD ECONOMIC OUTLOOK: Navigating Global Divergences

Figure 3.1. Fragmentation Keywords in Earnings Calls

of markets for energy, agricultural, and mineral

(Indices, 2013-15 = 100)

commodities could affect economies and global public

goods—namely, the energy transition. It focuses on the

1,600

400

Fragmentation keyword index, all sectors

following questions:

Fragmentation keyword index, commodity sector

1,400

350

Geopolitical risk index (right scale)

••

What makes commodity markets vulnerable in the

event of fragmentation?

1,200

300

••

Is there fragmentation in commodity markets, and if

1,000

250

so, what form does it take?

800

200

••

Which commodities are most vulnerable to disrup-

tions in international trade?

600

150

••

What would be the economic impact of commodity

400

100

market fragmentation across blocs and countries, as

200

50

well as on the global economy?

••

What might be the implications of such fragmenta-

0

0

2013

15

17

19

21

23:Q2

tion for the clean energy transition?

Sources: Caldara and Iacoviello (2022); Hassan and others (2019); NL Analytics,

Te chapter covers nearly all countries and focuses

Inc.; and IMF staff calculations.

Note: Fragmentation indices measure the average number of sentences, per

on 48 commodities, including agricultural goods,

thousand earnings calls, that mention at least one of the following keywords:

energy commodities—namely, coal, crude oil, and

deglobalization, reshoring, onshoring, nearshoring, friend-shoring, localization,

regionalization.

natural gas—and other mineral commodities. It

builds a unique database of commodity output, use,

and bilateral trade, and employs a combination of

Tere is little consensus on the economic costs of

descriptive statistics, empirical analysis, and model

fragmentation in the fast-growing literature. Esti-

simulations.

mates of long-term output losses from restricting

Te chapter simulates a highly stylized risk scenario,

the international flow of goods and services, finance,

in which commodity trade between two geopolitical

and technology range from 0.2 percent to 12 per-

blocs is persistently disrupted, to illustrate the chan-

cent of global GDP, depending on the scenario and

nels through which commodity market fragmentation

assumptions.³ Commodity markets could be another

could affect prices and output. Te main scenario

important channel through which further disruptions

defines the two theoretical blocs by using the 2022

in trade affect activity. Commodity production is hard

United Nations (UN) vote on the war in Ukraine as a

to relocate, as it is linked to natural endowments such

transparent starting point. However, the chapter exam-

as geological deposits or soil quality. Consumption of

ines alternative scenarios, including the role of neutral

commodities is often difficult to substitute in the short

countries and the impact of countries’ switching blocs,

term. Moreover, many commodities are crucial inputs

given the sensitivity of the analysis to bloc configura-

for manufacturing and technologies, including those

tions and the difficulty of assessing bloc configurations’

related to the clean energy transition.

plausibility.⁴

Against this backdrop, the chapter studies the

main channels through which further fragmentation

⁴Countries’ geopolitical alignment could be partly driven by trade

linkages and risk management strategies to reduce the fallout from

³See Aiyar and others (2023) for an overview of potential channels

spikes in geopolitical tensions. However, the endogenous formation

of impact. Recent studies that quantify aggregate losses from restricting

of blocs remains beyond the scope of the chapter. Te two-bloc

trade include Albrizio and others (2023); Attinasi, Boeckelmann, and

scenario presented here is meant to provide a clearly defined baseline

Meunier (2023); Bolhuis, Chen, and Kett (2023); Fally and Sayre

and to make the exercise comparable to those in the recent literature.

(2018); Felbermayr, Mahlkow, and Sandkamp (2022); Hakobyan,

Introducing neutral countries reduces the impact of fragmentation,

Meleshchuk, and Zymek (2023); and Javorcik and others (2022).

as discussed later in the chapter.

Chapter 4 of the April 2023 World Economic Outlook examines the

Online Annex 3.1 provides details on the commodities and coun-

consequences of restrictions on investment, and Chapter 3 of the April

tries and their allocation across blocs as well as data sources. Online

2023 Global Financial Stability Report does the same for portfolio flows,

Annex 3.5.2 discusses the robustness of some of the key findings to

whereas Cerdeiro and others (2021) and Góes and Bekkers (2022)

different bloc configurations. All online annexes are available at

quantify the losses once technological decoupling is also considered.

www.imf.org/en/Publications/WEO.

72

International Monetary Fund | October 2023

CHAPTER 3 Fragmentation and Commodity Markets: Vulnerabilities and Risks

Figure 3.2. Commodities: Key Characteristics

1. Average Share of Top Three Countries in World Production across

2. Distribution of Price Elasticities of Commodities’ Supply and

Commodities

Demand

(Percent of global production)

(Percent)

1.6

Agriculture

Short term

Long term

Energy

1.4

Minerals (mined)

1.2

Minerals (refined)

1.0

Wheat

0.8

Crude oil

0.6

Copper

Green

0.4

Nickel

transition

Cobalt

0.2

metals

Lithium

0.0

0

20

40

60

80

100

Demand

Supply

5

3. Upstreamness in Value Chains

4. Share of Traded World Production

100

(Index)

(Percent of global production)

4

80

World trade/

gross output

3

60

2

40

1

20

0

0

Minerals

Energy

Agriculture

Manufacturing Services

Agriculture

Energy

Minerals (mined) Minerals (refined)

60

5. Share of Countries that Import from Only One, Two, or Three

6. Vulnerability to Food Insecurity: Wheat

120

Suppliers

(Percent of annual consumption)

50

(Percent)

100

One supplier

Net imports

Domestic storage

40

80

Two suppliers

30

Three suppliers

60

20

40

10

20

0

0

Agriculture

Energy

Minerals (mined) Minerals (refined)

LICs

EMs

AEs

Sources: Antràs and others (2012); British Geological Survey; Dahl (2020); Fally and Sayre (2018); Food and Agriculture Organization of the United Nations; Gaulier

and Zignago (2010); International Energy Agency; US Department of Agriculture; US Geological Survey; and IMF staff calculations.

Note: “Energy” refers to coal, natural gas, and crude oil. This figure uses 2019 data due to data availability and to avoid biases caused by the pandemic. Panel 1

provides the share of global production that the top three producing countries account for (see Online Annex Figure 3.2.2) and gives averages across commodity

types. In panels 2 and 4, the horizontal lines in the bars represent the median, the squares the average, the bars the interquartile range, and the whiskers the

minimum and maximum values across commodities in the group. In panel 3, sectoral upstreamness is based on Antràs and others (2012) and is computed as the

weighted average position of an industry’s output in the value chain. The upstreamness index captures how far a specific sector is from the final end usage, with a

lower index value (minimum value of 1) implying that the sector is closer to final demand. Panel 4 does not include palladium and platinum due to data quality.

Panel 5 depicts the simple average across commodities of each group. Panel 6 depicts the simple average across countries within each income group, for 2019.

AEs = advanced economies; EMs = emerging markets; LICs = low-income countries.

Import dependence in agricultural commodities

populations to large swings in prices or food short-

can lead to food insecurity in case of trade disrup-

ages (Figure 3.2, panel 6). Te ramifications of food

tions, particularly in low-income countries. For

commodity shocks, which have been linked to social

instance, the average low-income country imports

unrest, conflict, and migration (Kelley and others

more than 80 percent of the wheat it consumes.

2015; Missirian and Schlenker 2017; Burke and

Given low storage capacity, consumption smooth-

McGuirk 2020), go beyond the economic analysis

ing can be difficult in these countries, exposing

that follows.

International Monetary Fund | October 2023

75

WORLD ECONOMIC OUTLOOK: Navigating Global Divergences

Figure 3.3. Commodity Trade and Distance of Military

associated with a decrease in trade in energy commodities

Alliances

by about 15 percent but it is associated with a more than

(Coefficients)

35 percent decline in minerals trade. Te exercise suggests

that changes in military alliances because of rising geopo-

0.0

litical tensions could go hand in hand with disruptions of

trade flows and fuel fragmentation of commodity trade.

-0.1

Fragmentation in Commodity Markets

-0.2

Tis section takes stock of various measures of

fragmentation. Te number of new interventions in

commodity trade has risen every year since 2018, with

-0.3

the increase initially fueled by tensions between China

and the United States and the COVID-19 pandemic.

In 2022, Russia’s war in Ukraine caused a major spike

-0.4

in new trade restrictions for commodities: there were

more than six times more new restrictions affecting

trade in commodities in 2022 than the 2016-19 aver-

-0.5

All commodities

Agriculture

Energy

Minerals

age. In contrast, trade-restricting measures on overall

trade increased 3.5 times (Figure 3.4, panel 1).¹⁰

Sources: Food and Agriculture Organization of the United Nations; Gaulier and

Zignago (2010); Leeds and others (2002); Signorino and Ritter (1999); and IMF

Price dispersion across locations can also be a symp-

staff calculations.

tom of fragmentation: as commodities are homoge-

Note: “Energy” refers to coal, natural gas, and crude oil. The bars in the figure

neous goods, they should trade under one price after

denote the point estimates, and the vertical lines represent 95 percent confidence

intervals. Standard errors are clustered at the importer country level. Estimates

transportation costs are accounted for. However, price

are based on an inverse hyperbolic sine specification to account for zeros.

dispersion increased in major commodity markets in

Distance of military alliances is measured using the similarity between countries’

portfolios of military alliances and standardized so its standard deviation is 1 in

2022, especially in those for some minerals, such as

each year. A standardized military distance of 1 is approximately the distance

lithium, and energy commodities (Figure 3.4, panel 2).

between India and Morocco in 2018.

For example, Russian coal traded at a price almost

three times lower than Australian coal in September

Sensitivity to Geopolitics

2022. Price dispersion for crude oil and natural gas

Analysis of trade patterns suggests that commodity

also rose as the war in Ukraine and associated sanc-

trade is historically associated with countries’ geopolitical

tions disrupted trade. Box 3.1 documents shifts in

alignment. Gravity equations, estimated for the commodi-

trade flows using real-time vessel-tracking data.

ties in the sample and augmented to include the similarity

Other proxies for fragmentation are changes in the

between countries’ portfolios of military alliances, show

number of FDI projects and cross-border mergers and

that bilateral commodity trade flows are negatively associ-

acquisitions, which are also indicators of future trade.

ated with military distance (Figure 3.3).⁹ However, notable

Tey were declining in the energy and mineral sectors

differences are apparent in the strength of the relationship

even before the war in Ukraine, which could presage a

across types of commodities: a one-standard-deviation

slowdown in commodity trade (Figure 3.4, panel 3).¹¹

increase in the distance of military alliances (approximately

Shifts have also occurred in the origin and destination

the distance between India and Morocco in 2018) is

¹⁰Trade interventions in the Global Trade Alert database, the

⁹Te gravity model is widely used to explain bilateral trade flows

source for the data in Figure 3.4, panel 1, include both measures

based on country and country pair characteristics that capture

that increase fragmentation, such as import tariffs and export restric-

trading costs, such as geographic distance or a common border, lan-

tions, and measures that aim to limit the economic fallout from

guage, or currency. Online Annex 3.3 provides details and robustness

fragmentation by encouraging diversification, such as subsidies for

checks. Distance in military alliances is associated with lower trade in

local producers, local-content requirements, and the like—although

minerals across specifications. Te results are more sensitive for other

a strict distinction between the two is difficult.

measures of geopolitical alignments, namely, the ideal point distance

¹¹Following Chapter 4 of the April 2023 World Economic Outlook,

based on UN votes, used in Chapter 4 of the April 2023 World

the analysis focuses on the number rather than the value for FDI

Economic Outlook in a similar analysis for FDI flows (see also Jaku-

and cross-border mergers and acquisitions. Data on values are

bik and Ruta 2023). Hakobyan, Meleshchuk, and Zymek (2023)

limited and often estimated. However, FDI values suggest a similar

examine distance in military alliances and sectoral trade flows.

decline in the commodity sector.

76

International Monetary Fund | October 2023

CHAPTER 3

Fragmentation and Commodity Markets: Vulnerabilities and Risks

Figure 3.4. Signs of Fragmentation

No measure of fragmentation is perfect. It is still

too early to assess how long-lasting price dispersion

700

1. Number of Trade Interventions by Sector

will be. Te decline in FDI could reflect moder-

(Indices, 2016-19 = 100)

600

ation in commodity prices since 2015, following

Agriculture

500

Energy

the decade-long commodity boom, and it is not

Minerals

clear, on account of lagging data, to what extent

400

All goods

trade-restricting measures have affected trade flows

300

(Goldberg and Reed 2023). However, taken together,

200

these measures suggest rising commodity market

100

fragmentation.

0

2009

10

11

12

13

14

15

16

17

18

19

20

21

22

Which Commodities Are Most Vulnerable?

400

2. Price Dispersion

(Difference between maximum and minimum as percentage of

To assess individual commodities’ vulnerability

minimum price across regions)

300

in the event of fragmentation, this section presents

Crude oil

Lithium

results from a single-commodity, multicountry partial

Coal

equilibrium model based on Alvarez and others (2023)

200

Wheat

Phosphate

(see also Online Annex 3.4 for details). It computes

price changes that would occur if trade for each of the

100

48 commodities included in the analysis were banned

across two blocs.

0

Jan.

Jul.

Jan.

Jul.

Jan.

Jul.

Jan.

Jul.

Jan.

Aug.

For illustrative purposes, the main scenario

2019

19

20

20

21

21

22

22

23

23

assumes that the two theoretical blocs are constructed

based on the 2022 UN vote on Russia’s war in

1,500

3. Foreign Direct Investment and Cross-Border Mergers

36,000

and Acquisitions

Agriculture

Ukraine. Te bloc comprising countries that voted

(Number of projects a year)

Energy

for Russia to withdraw from Ukraine is labeled the

Minerals

1,000

All sectors (right scale)

24,000

“US-Europe+ bloc”; the remaining countries are in

the “China-Russia+ bloc.”¹³ Te exercise assumes,

in a highly stylized and extreme way, that there is

500

12,000

no trade in a particular commodity between blocs,

whereas intrabloc trade of the commodity is unaf-

fected. Box 3.2 discusses more fluid experiences of

0

0

2003

05

07

09

11

13

15

17

19

21 22

fragmentation; investigating the impacts of those

intermediate scenarios is beyond the scope of the

Sources: Argus Media, Inc.; Bloomberg Finance L.P.; FDI Markets; Global Trade

chapter. Rather, the chapter’s goal is to identify rela-

Alert database; Refinitiv Eikon; UN Comtrade; and IMF staff calculations.

Note: Policy interventions are adjusted for reporting lags, and trade-liberalizing

tive vulnerabilities across commodity markets and to

interventions are excluded from calculations. Prices for crude oil, coal, and lithium

illustrate transmission channels, with the recognition

are market prices in different regional markets as retrieved from Bloomberg

Finance L.P. Wheat and phosphate price dispersion is estimated based on export

that partial interactions between blocs and arbi-

prices for countries that account for more than 5 percent of global exports, based

trage opportunities could mute the economic effects

on export data from UN Comtrade. Panel 3 presents the total number of foreign

implied by the model simulations.

direct investment and cross-border mergers and acquisitions projects at the global

level. The bars provide a breakdown by commodity group.

For each commodity, the model’s initial calibra-

tion is based on observed 2019 trade flows. Tey are

of commodity FDI and cross-border mergers and

assumed to reflect an integrated world, where goods

acquisitions. US and EU investors are increasingly tar-

are traded at one global price.¹⁴ Te trade ban across

geting projects in advanced economies, whereas China

and Russia have increased their investments in Africa

¹³See also Chapter 3 of the October 2022 Regional Economic

(see Online Annex Figure 3.2.5).¹²

Outlook: Asia and Pacific. More details on the countries in each bloc

and sensitivity checks for other bloc configurations are in Alvarez

and others (2023) and Online Annexes 3.1.2 and 3.5.2.

¹²Chapter 4 of the April 2023 World Economic Outlook documents

¹⁴Te assumption of perfect trade integration in the baseline over-

FDI flows are increasingly concentrated among geopolitically aligned

simplifies reality, as markets for some commodities were not perfectly

countries, particularly in strategic sectors.

integrated globally even before the war in Ukraine.

International Monetary Fund | October 2023

77

WORLD ECONOMIC OUTLOOK: Navigating Global Divergences

blocs yields bloc-specific prices that clear bloc-level

Figure 3.5. Price Changes Due to Fragmentation in Individual

supply and demand.

Commodity Markets

(Percent)

Fragmentation would induce opposite price effects

across blocs. Te price of a commodity falls in the bloc

Above

Palm oil,

that used to be a net exporter of that commodity and

500

soybean

increases in the net importing bloc. Te size of price

400

Cobalt,

copper,

changes depends on (1) bloc-level supply-and-demand

lithium,

300

imbalances prior to fragmentation—that is, the extent

US-Europe+

nickel

China-Russia+

to which a bloc relies on imports to satisfy its demand

200

at the integrated world price—and (2) the capacity of

100

demand and supply to respond to changing prices (the

price elasticities of demand and supply). Commodities

0

with inelastic demand and supply and with high imbal-

ances across blocs are more vulnerable to price changes.¹⁵

-100

Model simulations suggest that the potential price

-200

impact of fragmentation would vary significantly across

Agriculture

Energy

Minerals (mined) Minerals (refined)

commodities, with some potentially experiencing very

Sources: British Geological Survey; Food and Agriculture Organization of the United

large price increases (Figure 3.5; see Online Annex

Nations; Gaulier and Zignago (2010); International Energy Agency; United States

Figure 3.5.1 for the underlying commodity-specific

Geological Survey; and IMF staff calculations.

Note: Price effects are capped at 500 percent for readability. “Energy” refers to

results).¹⁶ In the China-Russia+ bloc, the price of

coal, natural gas, and crude oil. The black squares in the bars represent the

mined minerals, including cobalt, lithium, copper,

median; the bars, the interquartile range; and the whiskers, the data points within

1.5 times the interquartile range from the 25th or 75th percentile across

and nickel, which are critical for the green transition,

commodities in the group. The dots indicate outliers. Selected commodities which

would rise substantially. Production of these miner-

experience price increases higher than 500 percent are labeled. For the underlying

complete information on commodity-specific price changes, see Online Annex

als would be concentrated in a handful of countries

Figure 3.5.1. The bloc including the countries that voted for Russia’s withdrawal

in the US-Europe+ bloc, but they are largely used as

from Ukraine in the 2022 UN vote is labeled the “US-Europe+ bloc,” and the

inputs in the China-Russia+ bloc (see Online Annex

remaining countries are included in the “China-Russia+ bloc.”

Figure 3.2.6). At the same time, the US-Europe+ bloc

could experience similar increases in the prices of

country compositions of the two hypothetical blocs,

refined minerals, which are processed mostly in China,

described in Online Annex 3.5.2, suggest that, in a way

Russia, and South Africa.

similar to what occurs in the main simulation, fragmenta-

In contrast, the potential price changes for energy

tion would lead to significant price increases for minerals

and most agricultural commodities are more subdued

at the mining stage and for key agricultural staples (such

in the main simulation. Since the production of these

as soybeans) in the China-Russia+ bloc. In an alternative

commodities is less geographically concentrated,

bloc scenario, in which all emerging market and devel-

supply and demand are more balanced across the two

oping economies, excluding India, Indonesia, and Latin

blocs. However, palm oil and soybean represent two

American countries, are assigned to the China-Russia+

important outliers: more than 80 percent of produc-

bloc, the US-Europe+ bloc could experience large price

tion would occur in the US-Europe+ bloc, whereas

increases for some minerals. Tis is because key producers

most of the consumption would take place in the

are allocated to the other bloc. It could also become more

China-Russia+ bloc.

vulnerable in case of trade restrictions on some agricul-

Because of high geographic concentration, the

tural commodities (such as cocoa) and crude oil.

vulnerability of commodities in the event of fragmenta-

tion depends on the distribution of key exporters (and

importers) across blocs. Simulations based on different

Higher Commodity Price Volatility

Fragmented commodity markets would lead to

¹⁵Te exercise does not explicitly model storage, which is an

higher price volatility (see Jacks, O’Rourke, and

important feature of volatility smoothing. See among others, Williams

Williamson 2011, for historical evidence). Tis would

(1936), Gustafson (1958), and Wright and Williams (1982). Carter,

challenge public finances and fiscal and monetary

Rausser, and Smith (2011) provide a literature review.

frameworks, giving rise to potential procyclicality of

¹⁶Te following partial equilibrium results are based on Alvarez

and others (2023).

fiscal and monetary policies and hurting economic

78

International Monetary Fund | October 2023

CHAPTER 3 Fragmentation and Commodity Markets: Vulnerabilities and Risks

stability (Cavalcanti, Mohaddes, and Raissi 2015;

Figure 3.6. Wheat Price Increase in the US-Europe+ Bloc due

IMF 2023). Fragmentation can affect price volatility

to a Harvest Shock

(Percent)

through at least two channels: smaller market sizes and

countries switching blocs.¹⁷

10

Smaller Market Sizes

8

In a fragmented world, markets would become

smaller and bloc-level prices more responsive to

6

country-level shocks (see also Albrizio and others

2023). In the partial equilibrium model, the price

response is proportional to the supply shock’s size rela-

4

tive to the overall market. Tus, by restricting the set of

countries with which they trade, countries would face

2

larger price increases in response to the same negative

supply shocks.¹⁸ In an illustrative example, Figure 3.6

0

compares the price impact of a three-standard-deviation

Integrated world

Fragmented world

shock to the US wheat harvest in an integrated market

Sources: Food and Agriculture Organization of the United Nations; and IMF staff

with that in a fragmented market.¹⁹ Te same supply

calculations.

shock doubles the impact on wheat prices when trade

Note: The bars in the figure depict the change in the price of wheat in the

US-Europe+ bloc (those countries that voted for Russia to withdraw from Ukraine

is fragmented into two smaller blocs. Tis is important,

in the 2022 UN vote on the Ukraine war) from a three-standard-deviation negative

as climate change is expected to raise the variability of

shock to US wheat production. The figure compares the price increases in the bloc

in a free-trade world to those in a fragmented world.

agricultural output. A fragmented world, in which the

price response to supply shocks is amplified, would be

less able to cope with this challenge.

To illustrate price sensitivity to countries switching

blocs, Figure 3.7 shows the distribution of the great-

Countries Switching Blocs

est price increases each commodity can experience in

In a fragmented world, major commodity producers

a bloc when a single exporting country switches its

would face powerful incentives to switch geopolit-

alliance.²⁰ Minerals at the mining stage tend to be

ical allegiances, with such switching representing a

the most sensitive, given their highly concentrated

new source of supply shocks and price volatility. For

production. For example, South Africa produces

highly concentrated commodity markets, a single

one-third of the world’s manganese, a metal used in

exporting country switching to the other bloc could

steelmaking and batteries. If South Africa switched to

lead to a large supply gap and trigger hefty price

the US-Europe+ bloc, the price of manganese in the

changes. Uncertainty about a country’s geopoliti-

China-Russia+ bloc could rise more than 800 percent.

cal alignment could itself lead to price volatility as

traders update their priors regarding potentially large

Economic Impacts of Commodity Market

fragmentation-induced price swings.

Fragmentation

Tis section sheds light on the macroeconomic

¹⁷Other channels could include the impact on the financial

impacts of fragmenting commodity markets on

ecosystem linked to commodities, such as derivatives and insur-

individual economies, blocs, and the global economy.

ance (FSB 2023).

¹⁸In the single-commodity model, the price change in response to

Tree complementary analytical approaches are used.²¹

_

 

 

 

 

 

,

with Elasticity of Demand < 0 (Alvarez and others 2023).

²⁰Tese results are based on Alvarez and others (2023). Online

¹⁹Te United States accounts for about 7 percent of global and

Annex Figure 3.5.2 zooms into the results in Figure 3.7 by showing

15 percent of US-Europe+ bloc wheat production. A three-standard-­

the 15 commodities whose prices are most vulnerable to a single

deviation US harvest shock corresponds to about 60 percent of US

exporter switching blocs and the implied price changes.

wheat production, or 4 percent of global output, with wheat prices held

²¹Online Annexes 3.4-3.6; Alvarez and others (2023); and Bolhuis,

constant. Te exercise uses a price elasticity of supply of 0.2 and a price

Chen, and Kett (2023) discuss the assumptions, calibration, and

elasticity of demand of -0.85 (see Alvarez and others 2023). Lower

additional results of each model. None of the approaches consider the

elasticities would lead to higher price impacts, and fragmentation would

impact of fragmentation on productivity and innovation. Te role of

still double the price impact in this example.

the financial sector is also outside of the scope of the chapter.

International Monetary Fund | October 2023

79

WORLD ECONOMIC OUTLOOK: Navigating Global Divergences

Figure 3.7. Largest Price Increases Induced by a Single

fragmenting all commodity trade and to examine

Exporter Switching Blocs

the role of neutral blocs (see Box 3.3). Finally, the

(Percent)

dynamic effects on GDP and inflation are exam-

ined in a multiregion dynamic stochastic general

Above

Cobalt,

800

iron ore,

equilibrium model that includes energy and criti-

Cocoa Cocoa

Manganese

fluorspar

Platinum

cal minerals.

600

US-Europe+

Evidence from the Partial Equilibrium Model

400

China-Russia+

Several findings emerge from the partial equilibrium

approach. First, inefficiencies associated with restrict-

ing trade result in losses in bloc-level total surplus: the

200

global economy is worse off from the fragmentation

of trade in individual commodities (see Figure 3.8,

panel 1).²²

0

Agriculture

Energy

Minerals (mined) Minerals (refined)

Second, bloc-level changes in total surplus are gen-

erally small (with some notable exceptions²³), masking

Sources: British Geological Survey; Food and Agriculture Organization of the United

important heterogeneities across countries. Within

Nations; Gaulier and Zignago (2010); International Energy Agency; United States

Geological Survey; and IMF staff calculations.

each bloc, some countries would experience an increase

Note: Price effects are capped at 800 percent in the figure for readability.

in surplus (net-exporting countries in a net-importing

“Energy” refers to coal, natural gas, and crude oil. Each observation in the box

plots represents the largest price increase that a commodity can experience in

bloc and net-importing countries in a net-exporting

each bloc from a single exporting country’s switching to the other bloc. Note also

bloc), and some experience a decline. Such changes

that the US (China) is not allowed to switch away from the US-Europe+

(China-Russia+) bloc. The black squares in the bars represent the median; the

would be small for most countries as a share of gross

bars, the interquartile range; and the whiskers, the data points within 1.5 times

national expenditure but could be very sizable for a

the interquartile range from the 25th or 75th percentile across commodities in the

group. The dots indicate outliers; the commodities representing the largest outliers

few commodity importers and exporters, as shown in

are labeled. For the underlying complete information on commodity-specific price

Online Annex Figure 3.5.4. For instance, fragmenta-

changes, see Online Annex Figure 3.5.2. The bloc including the countries that

tion of copper at the mining stage would reduce sur-

voted for Russia’s withdrawal from Ukraine in the 2022 UN vote is labeled the

“US-Europe+ bloc,” and the remaining countries are included in the

plus by as much as 2.5 to 5 percent of gross national

“China-Russia+ bloc.”

expenditure in Chile and Peru, both exporters of

copper to the US-Europe+ bloc, in which prices would

fall. At the same time, it would lead to large surplus

First, the partial equilibrium model discussed earlier

gains in Kazakhstan and Mongolia, which would

is leveraged to compute changes in producer and

scale up exports at higher prices to the copper-scarce

consumer surplus due to fragmentation in individual

China-Russia+ bloc (Figure 3.8, panel 2).

commodity markets. Te resulting change in total sur-

Tird, restricting trade in commodities that are

plus is used as an indicator of economic impact. Tis

less price-vulnerable could still generate sizable

approach identifies the most macro-relevant commodi-

ties. It accounts for the changes in price and quantities

²²Tis result and the following are based Alvarez and others

consumed or produced of each commodity because of

(2023). Tey also provide the analytical proof. In an integrated

fragmentation. However, due to its partial equilibrium

world, trade patterns reflect the efficient allocation of resources glob-

ally, with countries specializing in commodities for which they have

nature, the approach does not account for sectoral

comparative advantage (cost-effective deposits or suitable climate

spillover effects, nor does it allow for the simultane-

conditions). After fragmentation, trade patterns no longer reflect

ous disruption of trade in many commodities, which

these comparative advantages.

could have opposing or reinforcing effects within the

²³Online Annex Figure 3.5.3 shows the five largest surplus losses

at the bloc level from the fragmentation of a single commodity.

same country.

Such data points are marked as outliers in Figure 3.8, panel 1,

Two general equilibrium models in the chapter

capped at -0.05 percent of gross national expenditure. In the

overcome these shortcomings. A static multicoun-

main simulation, trade fragmentation of palm oil or copper at the

mining stage could lead to surplus losses in the China-Russia+ bloc

try, multisector trade model, which accounts for

of more than 1 percent of gross national expenditure, and trade

all input-output linkages across sectors, is used to

fragmentation of iron ore or soybeans to surplus losses of more

simulate the long-term GDP losses associated with

than 0.5 percent.

80

International Monetary Fund | October 2023

CHAPTER 3 Fragmentation and Commodity Markets: Vulnerabilities and Risks

Figure 3.8. Surplus Changes due to Fragmentation in

Finally, surplus declines would generally be larger

Individual Commodity Markets

in the hypothetical China-Russia+ bloc. Commodities

that are most vulnerable are more broadly consumed in

China-Russia+

US-Europe+

this bloc (Online Annex Figure 3.5.3).²⁴

0.01

1. Surplus Changes by Bloc and Commodity Group

(Percent of bloc-level GNE)

0.00

Evidence from the Trade Model

-0.01

Te general equilibrium multicountry, multisector

trade model presented in Box 3.3 simulates long-term

-0.02

GDP effects from the disruption of all commodity

-0.03

Copper,

trade. Broad differences are seen in the impact across

Palm oil,

iron ore,

–0.04

soybean

Iron ore

cobalt

countries, with some experiencing sizable losses.

Low-income countries could suffer deeper losses, on

Below

-0.05

Agriculture

Energy

Minerals (mined) Minerals (refined)

average estimated at 1.2 percent, given their high

dependence on agricultural trade. For some of these

2. Surplus Changes for Top Two Net Exporters in Each Bloc for

countries losses could amount to more than 2 per-

Selected Commodities

(Percent of GNE)

cent of GDP. Consistent with the single-commodity

Above

exercise, the hypothetical China-Russia+ bloc is more

20

affected by fragmentation, yet the global GDP loss, at

10

roughly 0.3 percent, is modest as a result of offsetting

effects across countries.²⁵

0

Te economic impact can be greatly reduced if

commodity trade is only partially restricted. Illustrative

-10

simulations, in which countries that abstained from the

UN vote on Ukraine are assumed to trade commodi-

-20

ties freely, point to much smaller effects of trade bar-

riers between the US-Europe+ and the China-Russia+

Copper

Crude oil

Palm oil

blocs. Long-term changes in global GDP from this

scenario would be negligible, with meaningful losses

Sources: British Geological Survey; Food and Agriculture Organization of the United

Nations; Gaulier and Zignago (2010); International Energy Agency; United States

only in Russia. Tis is in line with historical evidence

Geological Survey; and IMF staff calculations.

on the ability of demand and supply of commodities

Note: “Energy” refers to coal, natural gas, and crude oil. In panel 1, each data

to adjust in response to trade restrictions (Box 3.2).

point in the box plots represents the total bloc-level surplus change from

fragmenting trade in a single commodity. The black squares in the bars represent

the median, the bars are the interquartile range, and the whiskers reflect the data

points within 1.5 times the interquartile range from the 25th or 75th percentile

Evidence from the Dynamic Macroeconomic Model

across commodities in the group. Dots indicate outliers; the commodities

associated with the largest surplus declines are labeled. The bloc including the

Tis subsection uses a dynamic stochastic general

countries that voted for Russia’s withdrawal from Ukraine in the 2022 UN vote is

labeled the “US-Europe+ bloc,” and the remaining countries are included in the

equilibrium framework to assess the dynamic GDP

“China-Russia+ bloc.” Data labels in the figure use International Organization for

and inflation effects of commodity fragmentation.

Standardization (ISO) country codes. GNE = gross national expenditure.

Te model is based on an augmented version of the

²⁴Sensitivity checks in Online Annex 3.5.2 show that this holds

for a bloc configuration based on existing trade relationships. Alter-

surplus declines. For example, energy commodities

natively, if all emerging market and developing economies, excluding

are not particularly vulnerable under the baseline bloc

India, Indonesia, and Latin American countries, are assigned to the

configuration, but the associated declines in surplus

China-Russia+ bloc, the US-Europe+ bloc could experience larger

surplus losses, mainly on account of oil market disruptions.

would be more significant, because energy commod-

²⁵Global GDP losses from restricting commodity flows between

ities are widely consumed and produced. In contrast,

blocs constitute about 15 percent of the loss from restricting all

minerals could experience strong price changes, but

trade. In comparison, commodities represent only 10 percent of total

trade. Te larger losses from fragmenting energy and agricultural

the surplus impact would be more subdued, given

markets in Bolhuis, Chen, and Kett (2023) stem from the assump-

their (so far) more limited relevance in most countries’

tion of full autarky compared with the no-trade-between-blocs

production and consumption.

scenario in the chapter.

International Monetary Fund | October 2023

81

WORLD ECONOMIC OUTLOOK: Navigating Global Divergences

IMF’s Global Macroeconomic Model for the Energy

Figure 3.9. Impact of Fragmentation on Real GDP and

Transition.²⁶ It includes the production, consumption,

Inflation

(Percent deviation from baseline)

and trade of energy from fossil and renewable sources

as well as four minerals critical to the energy transi-

Energy and minerals

Minerals

Natural gas

Crude oil

tion. Commodities include crude oil, coal, natural gas,

copper, nickel, cobalt, and lithium, capturing about

0.2

1. Average Deviation of GDP over First Three Years

70 percent of the value of global commodity trade.

0.1

Fragmentation is modeled as a ban on trading these

0.0

commodities between the two hypothetical blocs,

-0.1

which comprise six different regions.

-0.2

In the model, fragmentation affects activity through

several channels. First, the trade ban induces expenditure

-0.3

Europe

switching and trade diversion. Second, temporary imbal-

-0.4

ances between supply and demand within blocs emerge

-0.5

World

US-Europe+

China-Russia+

until prices adjust to clear markets. Such imbalances

generate swings in commodity prices. Finally, rigidities

1.2

2. Deviation of Inflation in Year One

affect the speed of adjustment of output, use, and trade,

1.0

as well as overall macroeconomic effects.

Europe

0.8

Te output and inflation effects could vary sig-

0.6

nificantly across regions, blocs, and commodities

0.4

(Figure 3.9). Comparison of the impact on individual

0.2

commodities highlights the channels at play. Te effects

0.0

of fragmenting trade in oil and gas would be quite dif-

-0.2

ferent, even though the distribution of oil and gas con-

-0.4

sumption and production would be similar across blocs.

World

US-Europe+

China-Russia+

For oil, countries could quickly switch to trading part-

Sources: British Geological Survey; Food and Agriculture Organization of the United

ners within their bloc, with limited impact on GDP. By

Nations; Gaulier and Zignago (2010); Global Macroeconomic Model for the Energy

contrast, rigidities such as the need for pipelines or other

Transition; Organisation for Economic Co-operation and Development,

structures would constrain natural gas trade diversion,

Inter-Country Input-Output Tables; United States Geological Survey; and IMF staff

calculations.

with more pronounced effects on GDP. GDP would

Note: “Energy” refers to coal, natural gas, and crude oil. Region-level results are

decline and inflation would increase in both blocs.

aggregated to the bloc and world levels using weights based on GDP at

purchasing power parity. The bloc including the countries that voted for Russia’s

In the case of minerals, simulations highlight the

withdrawal from Ukraine in the 2022 UN vote is labeled the “US-Europe+ bloc,”

importance of the geographic distribution of mining

and the remaining countries are included in the “China-Russia+ bloc.”

production and rigidities in scaling up refining capacity.

On the one hand, fragmentation could lead to a steep

rise in prices in the China-Russia+ bloc and sizable

up refining capacity. Tat bloc would also experience a

declines in real GDP. Roughly 80 percent of the supply

GDP decline from mineral market fragmentation.

of the four minerals is mined in the US-Europe+ bloc,

Trade fragmentation of all seven commodities would

and minerals are used intensively in the China-Russia+

be associated with a global GDP loss of about 0.3 per-

bloc’s sizable manufacturing and construction sector. On

cent. However, as in the partial equilibrium and trade

the other hand, the US-Europe+ bloc would not be able

models, sizable differences are observed across and

to benefit from the relative oversupply of minerals at the

within blocs. Te simulated losses would be larger in

mining stage because it would take several years to scale

the China-Russia+ bloc. Within the US-Europe+ bloc,

Europe could experience a sizable impact on inflation

²⁶Te model was first used in Chapter 3 of the October 2022

(as much as 100 basis points or more) and GDP, with

World Economic Outlook. It is augmented here by (1) including

that impact driven mainly by the fragmentation of oil

the possibility of segmenting tradable energy markets and (2)

explicitly modeling two types of mineral aggregates composed

and gas markets.

of copper and nickel as well as cobalt and lithium, respectively.

Several caveats are worth highlighting. Whereas

Te augmented model has six regions: the United States, the

the model provides regional granularity, it masks the

European Union, US-EU-leaning countries, China, Russia, and

China-Russia-leaning countries.

heterogeneity of effects across countries, given the

82

International Monetary Fund | October 2023

WORLD ECONOMIC OUTLOOK: Navigating Global Divergences

Figure 3.10. Impact of Fragmentation of Critical Mineral

highly concentrated and difficult-to-relocate produc-

Markets on Investment in Renewables and Electric Vehicles,

tion, hard-to-substitute consumption, and critical role

2030

as inputs for manufacturing and key technologies. Frag-

(Percent deviation from net-zero-emissions scenario without

mentation in commodity markets is on the rise. Mea-

fragmentation)

sures restricting commodity trade surged in 2022, price

20

differentials across geographic markets have widened

Investment in renewables

Production of EVs

for selected commodities, and FDI flows in commodity

sectors are in decline—the latter a trend that started

0

before the war in Ukraine.

Illustrative model simulations suggest that more severe

-20

fragmentation could cause large changes in commodity

prices, depending on the resulting supply-and-demand

-40

imbalances and commodities’ elasticities of supply and

demand. Critical minerals for the energy transition and

-60

some highly traded agricultural goods are highly vulner-

able in the event of fragmentation.

-80

A fragmented world would be more volatile. Com-

World

US-

China

World

US-

China-

modity price volatility could intensify as a result of

Europe+

Russia+

Europe+

Russia+

smaller market sizes and the incentives for producers

Sources: British Geological Survey; Gaulier and Zignago (2010); IMF, Global

to switch geopolitical allegiances. Tis could result in

Macroeconomic Model for the Energy Transition; International Energy Agency;

United States Geological Survey; and IMF staff estimates.

volatile inflation dynamics, making monetary policy

Note: The bars and dots in the figure report the change in real investment in

more complex.

renewable energy and the production of EVs in a fragmented world relative to the

Te potential impacts of fragmentation differ vastly

net-zero-emissions path, with demand for cobalt, copper, lithium, and nickel

increasing as projected by the International Energy Agency’s net-zero-emissions

across countries, with offsetting effects across con-

scenario (in an integrated world). Country-level variables are aggregated to the

sumer and producer countries resulting in modest

bloc and world levels using weights based on GDP at purchasing power parity in

the bars and on greenhouse gas emissions in the dots. The bloc including the

output losses at the global level. Low-income coun-

countries that voted for Russia’s withdrawal from Ukraine in the 2022 UN vote is

tries, on average, would experience significantly deeper

labeled the “US-Europe+ bloc,” and the remaining countries are included in the

“China-Russia+ bloc.” EVs = electric vehicles.

long-term output declines. Given the heavy reliance

on agricultural imports among many low-income

countries, fragmentation of agricultural commodities

would increase to about 30 percent if one uses green-

would raise important food security concerns. Illus-

house gas emissions to weigh the regional response

trative model simulations suggest that a hypothetical

of investment in renewables and EVs. Te measure

China-Russia+ bloc could be more affected economi-

accounts for the greater emissions intensity of activity

cally than a US-Europe+ bloc, although the economic

in the China-Russia+ bloc and hence the greater effort

impact would be reduced if commodity trade was only

needed to achieve global emissions mitigation goals.³⁰

partially restricted or there was a nonaligned bloc.

Decarbonizing the world economy would be more

Overall, further fragmentation of commodity markets

difficult if the market for minerals is fragmented.

could deliver an additional blow in an already challeng-

ing environment of slow global growth, tight financial

Summary and Policy Implications

conditions, and high debt, a blow that would be partic-

ularly harsh for some of the most vulnerable economies.

Commodity markets are an important channel

Fragmentation in critical mineral markets could

through which geopolitical fragmentation can affect

make the clean energy transition more costly, raising

the economy. Many features of commodities underpin

the risks of delaying necessary climate change mitiga-

their vulnerability in the event of fragmentation: their

tion. It could add to the upward price pressure in the

mineral-scarce bloc in the chapter’s illustrative model

³⁰Te China-Russia+ bloc accounted for more than half of

greenhouse gas emissions in 2020, but only a third of global

simulation. Te mineral-rich bloc in the simulation

GDP. Hence, global investment losses are significantly larger when

could not reap the benefits from oversupply in the near

bloc-level changes are aggregated using emissions (the yellow dots

term because it would be unable to scale up refining

in Figure 3.10) rather than purchasing-power-parity-weighted GDP

(the bars in Figure 3.10).

and processing capacity quickly. In the simulation,

84

International Monetary Fund | October 2023

CHAPTER 3 Fragmentation and Commodity Markets: Vulnerabilities and Risks

Box 3.1. Commodity Trade Tensions: Evidence from Tanker Traffic Data

Since its invasion of Ukraine, Russia’s oil exports have

and insurance services to tankers carrying Russian

been subject to sanctions and have been voluntarily

commodities above certain price thresholds.

shunned by firms. What has been the impact on oil

Automatic Identification System data reveal that the

trade flows? Granular real-time data on tanker ship-

traffic patterns of Russia’s tankers have since changed

ping patterns from the Automatic Identification System¹

substantially (Figure 3.1.1). Tanker shipments from Rus-

uncover significant shifts in routes, resulting in economic

sian ports to Japan, the United States, and the European

inefficiencies.

Union declined between April-June 2019 and the same

period in 2023. Other countries are also now providing

Te European Union, United Kingdom, and United

oil supplies. For example, the European Union receives

States banned most imports of crude oil and petro-

more shipments from countries such as Norway, the

leum products from Russia after Russia’s invasion of

United Arab Emirates, and the United States, but this

Ukraine. Western restrictions on dollar payments have

extends the length of tanker routes by 20 percent.²

been reported to be a barrier to shipments. Group of

On the flip side, Russian oil shipments rose after the

Seven (G7) members also prohibited transportation

invasion to countries such as China, India, Türkiye,

and the United Arab Emirates. About 35 to 40 per-

cent of India’s crude oil imports came from Russia

Te authors of this box are Seung Mo Choi and Alessandra Sozzi.

during April-June 2023, a stark rise from less than

¹Te Automatic Identification System is a mandatory

5 percent before the war in Ukraine. While India’s oil

self-reporting system for all ships above 300 gross tons. It has

exports (mostly petroleum products) are small relative

been used to construct real-time trade indicators (examples are

to its oil imports (mostly crude oil), India increased its

included in Arslanalp, Marini, and Tumbarello 2019; Cer-

oil exports to the European Union substantially.

deiro and others 2020; and Arslanalp, Koepke, and Verschuur

2021). PortWatch (https://www.imf.org/portwatch) is an online

platform that monitors trade disruptions and assesses spillovers

²UNCTAD (2022) documents a rise in tanker freight rates

through port-to-port links.

following the Russian invasion of Ukraine.

Figure 3.1.1. Changes in Tanker Shipments from Russia’s Ports from 2019:Q2 to 2023:Q2

(Metric tons, decreases in blue and increases in red)

-2.5M

-1.5M

-750K -100K

100K

750K

1.5M

2.5M

3.5M

Sources: Natural Earth; UN Global Platform; and IMF staff calculations.

Note: The bubble size indicates the magnitude of the change for the destination port. Lines indicate travel routes.

International Monetary Fund | October 2023

87

CHAPTER 3 Fragmentation and Commodity Markets: Vulnerabilities and Risks

Box 3.3. The Uneven Economic Effects of Commodity Market Fragmentation

Fragmentation of commodity markets affects countries

elasticity of substitution between commodities and

and households differently. This box demonstrates that

other inputs in the production of manufactured goods.

low-income countries are more vulnerable in the event

Trade costs are set such that there is no commodity

of fragmentation, especially of agricultural commodities,

trade between blocs.

owing to their greater reliance on food imports. The find-

Results show that the aggregate impact of commod-

ing raises important food security concerns should further

ity fragmentation would be moderate, with a global

fragmentation materialize.

GDP loss of 0.3 percent (Figure 3.3.1). However,

there would be large differences within and across

To quantify the impact on long-term GDP of

blocs. Some economies might benefit from trade diver-

fragmentating trade in multiple commodities simul-

sion as competitors lose access to export markets. Most

taneously, a multicountry, multisector trade model is

would experience permanent output declines. Losses

used in this box, following Caliendo and Parro (2015).

would be larger in countries where commodity trade

Bolhuis, Chen, and Kett (2023) augment the model

with the other bloc was significant. Te China-Russia+

to account for trade and production of 133 commod-

bloc and low-income countries—whose economies are

ities across 145 countries. Labor is the only factor of

more commodity-intensive—would lose more.

production, and productivity is exogenous. Commod-

Low-income countries’ high dependence on imports

ities are used as intermediate inputs, with a long-term

of agricultural goods would make them particularly

supply elasticity of 1. Te model accounts for the

vulnerable (Figure 3.3.2). Disrupting trade in food

input-output structure of global trade and assumes low

commodities alone would lead to losses of 1 percent of

GDP. Commodity fragmentation could also have high

social and humanitarian costs and would be partic-

Te authors of this box are Marijn Bolhuis, Jiaqian Chen,

ularly harmful for lower-income households, which

and Benjamin Kett. See Bolhuis, Chen, and Kett (2023) for

further details.

spend a large share of their incomes on food and fuel.

Figure 3.3.1. Estimated Output Losses

Figure 3.3.2. Estimated GDP Losses in

(Percent deviation from baseline)

Low-Income Countries and Others

(Percent deviation from baseline)

0.0

0.2

-0.1

0.0

-0.2

-0.2

-0.3

-0.4

-0.4

-0.6

-0.5

-0.8

All commodities

-1.0

Agriculture

-0.6

Energy

-1.2

Minerals

–0.7

Global

US-Europe+

China-Russia+

-1.4

All commodities

Low-income countries

Others

Sources: British Geological Survey; Eora Global Supply Chain

Sources: British Geological Survey; Eora Global Supply Chain

database; Food and Agriculture Organization of the United

database; Food and Agriculture Organization of the United

Nations; Gaulier and Zignago (2010); US Geological Survey;

Nations; Gaulier and Zignago (2010); US Geological Survey;

and IMF staff calculations.

and IMF staff calculations.

Note: The bars represent the losses in GDP relative to

Note: The bars represent the losses in GDP relative to

baseline from eliminating trade in commodities across

baseline from eliminating trade in groups of commodities

hypothetical blocs. Country-level losses are aggregated

across hypothetical blocs. Country-level losses are

using weights based on GDP at purchasing power parity. For

aggregated using weights based on GDP at purchasing

details, see Bolhuis, Chen, and Kett (2023).

power parity. For details, see Bolhuis, Chen, and Kett (2023).

International Monetary Fund | October 2023

89

Content      ..     1      2      3      4      ..