## 1. Structural Features of Global Trade and Financial Flows Network: Gatekeepers

## Source details

**Canonical URL:** [1. Structural Features of Global Trade and Financial Flows Network: Gatekeepers](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2012/_031512.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2012/_031512.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2012/_031512.pdf.json)

---

### Motivation and two-stage approach
- Interconnectedness delivers benefits (access to new markets, cheaper funding, more efficient allocation of capital) and risks (fast, wide shock transmission); fragility and limited policy buffers in advanced economy core raise systemic volatility.
- Two-part analytical approach:
  - Stage I — Architecture: clique percolation cluster (network) analysis to map overlapping clusters, a global core (mainly the systemic-5), and gatekeepers/connectors (e.g., China as gatekeeper to Asian supply chain).
  - Stage II — Applications: case studies to analyze how cluster functional/economic features shape shock propagation and policy implications; context-dependent outcomes (no general link between rising interconnectedness and output synchronicity).

### Key hypotheses and implications
- Gatekeepers can act as dampeners or amplifiers of shocks; sensitivity analyses identify gatekeepers as possible pressure points or “circuit breakers”.
- Situating an economy within its cluster(s) can:
  - facilitate sharper assessment of shock propagation and risks;
  - highlight which partners and which policies/buffers matter for systemic stability;
  - provide a basis for dialogue on macroeconomic, macro-prudential, and capital flow management policies within clusters.
- Caveat: cluster analysis generates hypotheses requiring confirmation through deeper economic analysis to avoid type I and type II errors.

### Framework — Stage I: Mapping the Network Architecture (behavior- or risk-neutral)
- Problem: dense bilateral trade and financial connections create a high-dimensional, visually opaque network.
- Methodology and data rules:
  - Clique percolation clustering condenses bilateral trade and financial data into overlapping subgroups; overlapping clusters preserve multi-affiliation.
  - Inclusion threshold: only bilateral connections above 15 percent of a country’s overall connections in that category are used.
  - Data vintage: all data are as of end-2009.
  - Aggregation: bilateral links summed across trade and financial dimensions with links weighted by their relative shares in respective datasets.
  - Gatekeeper identification: iterative—if removing an economy causes a cluster to drop out, that economy is identified as connecting that cluster to the network.
- Stylized architecture: overlapping clusters, a core (intersection of many clusters), and gatekeepers linking clusters and core.
- Stylized role hypotheses:
  - Core: shocks and policies in the core can have global spillovers.
  - Clusters: shocks may have larger initial impact within clusters given denser connections (Europe highlighted).
  - Gatekeepers: may transmit or dampen shocks (example: Austria’s fiscal recapitalization limited deleveraging and acted as a dampener in CESEE).

### Framework — Stage II: Analyzing Clusters and Evaluating Hypotheses
- Need to examine functional characteristics and economic context to test Stage I hypotheses.
- Interconnections can:
  - dissipate certain shocks (easier substitution);
  - amplify or propagate others (via gatekeepers).
- Durability: for enduring clusters, gatekeepers may be perceived to dissipate shocks; otherwise cluster membership may change.
- Existing literature ambiguous—effects depend on:
  - extent and nature of integration (inter- vs. intra-industry trade);
  - types of shocks (industry-level vs. common shocks);
  - levels of economic and financial development;
  - type of financial clustering and prudential frameworks.

### Policy considerations and potential actions
- Trade-related:
  - When close trade linkages produce synchronized business cycles, loss from foregoing monetary policy independence may be smaller (OCA logic), strengthening case for policy coordination.
  - Where cycles are not synchronized, policy flexibility is desirable.
  - Surveillance can reveal cluster-level cycle relationships and policy “tensions”.
- Finance-related:
  - Close financial integration can drive capital inflows and expansion in upturns and sudden stops in downturns.
  - Information sharing and coordination to manage inflows may be warranted; reactions to inflows in one country can divert flows elsewhere.
  - Coordination with source countries of inflows can improve policy effectiveness.

### Caveats and limits
- Results sensitive to data quality and quantity; more comprehensive financial exposure data would sharpen findings.
- Cluster analysis is a step toward understanding interconnectedness but is not definitive; hypotheses must be tested with detailed economic analysis and models.

---

### Data and methodology (detailed)
- Data sources: IMF Direction of Trade Statistics (aggregate), UN Comtrade (various commodities, capital goods, manufacturing goods), BIS consolidated and locational banking statistics, IMF CPIS, IMF CDIS.
- Inclusion threshold: 15 percent rule applied endogenously to maximize cluster overlaps.
- Aggregate network findings rely on weighted sums across datasets to avoid adding absolute numbers from differently scaled datasets.

### Aggregate architecture — key quantitative findings
- Broad features:
  - Large number of overlapping clusters.
  - Global core economies (mainly the systemic-5) connected to most other economies and many clusters.
  - Multiple gatekeepers in several clusters, implying shocks often must pass multiple “gates” to spread.
  - Some economies do not appear in individual clusters but remain part of the global system (may reflect limited data on some dimensions).
- Table 1 (aggregate network) headline counts:
  - total groups of gatekeepers: 21;
  - groups that cut across regions: 18;
  - groups that fit regional categories: 3;
  - United States identified as a gatekeeper to 16 groups;
  - China identified as a gatekeeper to 12 groups.
- Output correlations:
  - average real GDP growth correlation over the past 20 years is 0.37 within clusters versus 0.12 for the world.
  - Interpretation: within-cluster co-movement is three times higher than world average but not more than a standard deviation apart; clusters may often mitigate rather than amplify local shocks.

### Representative clusters and illustrative group-level implications
- Asian cluster around China and the United States; Euro Area cluster around Germany.
- Examples of gatekeeper effects:
  - Group P: problems in Greece can spill beyond its cluster because gatekeepers Italy and Germany are impacted; Italy’s stress could spread across Europe, the Middle East, Central Asia, and sub Saharan Africa.
  - Group O/N/J/U/K: Italy serves as a gatekeeper/conduit into CESEE, Belgium/France/Germany/Netherlands, and sub Saharan Africa; Hong Kong SAR may intermediate equity financing between Asian economies and U.S. investment via G-SIFIs.
- Surprises: some countries link to unexpected clusters (e.g., Lithuania closer to Poland than to Estonia/Latvia; Algeria/Egypt/Lebanon/Tunisia closer to European clusters).

### Durability over time
- Cluster analysis over 6 time periods from 2000 to 2009 shows broad stability; notable exception—China grew rapidly and became more central, concentrating the Asian supply chain.

### Sectoral/granular analysis and extensions
- Aggregated network can be decomposed into sub-networks (machinery, capital goods, particular financial exposures such as through G-SIFIs).
- For sector-concentrated countries (commodity exporters), granular cluster/gatekeeper analysis yields more detailed vulnerability insights.
- Augmentations suggested:
  - include membership in formal currency areas;
  - construct clusters based on correlated asset prices (similarities rather than direct exposures).

### Vulnerability / stress-testing application (sensitivity exercise)
- Method: impose exogenous/mechanical reductions of links and compare resulting clusters/gatekeepers with pre-perturbed system; endogenous minimum threshold determines which links prune.
- Illustrative uniform stress of 15 percentage points to all trade and financial links causes some links to fall below threshold.
- Interpretation rules (as in Figure 7):
  - each country associated with two numbers: left = share of envisioned shock transmitted to it (incoming); right = share of its effects on others (outgoing).
  - Italy identified as a “pressure point” whose shock significantly affects multiple regions; if Italy has buffers it can act as a “circuit breaker”.
  - United Kingdom acts as shock absorber from Europe and conduit to the United States due to role as global intermediator of U.S. dollar assets.
  - China’s density lies more within Asia; stickiness in trade relationships could mute some shocks, but stress in China could materially affect clusters including the United States.
- Caveat: mechanical exercise maps potential transmission routes and feedback loops; it is not a quantitative gauge of relative intensity—careful shock identification and robust empirical work are needed.

---

### Cluster case studies — selected highlights

- Asian vertically integrated supply chain (China, Hong Kong SAR, Indonesia, Japan, Korea, Malaysia, Philippines, Singapore, Thailand, Vietnam)
  - Structure: China a key gatekeeper; Hong Kong SAR and Singapore link cluster to U.S. and U.K. financial centers.
  - Cycles: close linkage to global electronics cycle; business cycles closely—but not entirely—synchronized with China’s.
  - Policy implications:
    - Exchange rate co-movement is high; downstream countries may need relative currency stability to preserve production positions, limiting exchange rate as an adjustment tool.
    - Macro-prudential and capital flow management policies can augment policy space when exchange rate flexibility is constrained.
  - Example: China’s July 2005 exchange rate reform led to cumulative renminbi appreciation of 20¼ percent vis-à-vis the U.S. dollar by end-2008.
  - Country responses to capital inflows (2003–04 surge):
    - Malaysia abandoned dollar peg coincident with China and managed exchange rate; fiscal consolidation allowed slower appreciation.
    - Thailand saw faster appreciation and imposed capital controls end-2006; measures partly reversed over 1½ years.
    - Philippines experienced largest nominal appreciation against both dollar and renminbi by 2008 due to fiscal dominance limiting monetary policy.

- Saudi Arabia — Commodity cluster
  - Gatekeepers: United States, China, Korea, and Japan.
  - Financial development: domestic financial deepening strengthened monetary transmission; U.S. dollar peg means U.S. monetary conditions influence Saudi non-oil sector.
  - Event (September 2007–October 2008):
    - U.S. Federal Reserve lowered policy rate from 5.25 percent to 1 percent.
    - Crude oil surged from over $65 per barrel in mid-2007 to over $130 per barrel by summer 2008.
    - Saudi policy rate cut from 5 percent in October 2007 to 2 percent by mid-2008 to prevent speculative inflows and maintain peg.
    - Annual credit growth rose from 6 percent in early 2007 to over 30 percent in July 2008.
    - Double digit inflation by mid-2008.
  - Trade-offs: counter-cyclical tensions between maintaining peg and managing domestic credit/demand; policy options include strong fiscal management and refined macro-prudential tools.

- Nordic-Baltic financial cluster
  - Composition: Nordic banks’ home countries (Denmark, Norway, Sweden) as gatekeepers; Baltic hosts (Estonia, Latvia, Lithuania).
  - Pre-crisis: parent banks transmitted global liquidity to hosts, fueling credit/asset booms; monetary policy channel clogged in hosts.
  - Crisis dynamics: funding withdrawal by Swedish banks amplified shock; reliance on Swedish policy space to back-stop Baltic exposures.
  - Swedish response: provided liquidity (kroner and FX swaps), supported banks, participated in Vienna Initiative; Latvia had IMF-supported program.
  - Evidence: structural VARs show Swedish monetary policy lowered spillovers to Baltic money markets and GDP, especially Estonia and Latvia.
  - Policy lessons:
    - Strengthen fiscal stabilizers and structural measures in Baltic hosts to limit credit booms and contingent liabilities.
    - Nordic home authorities should monitor and curb intra-group exposures via information sharing and coordination.
    - Gatekeeper actions can initially amplify but, via buffers, act as circuit-breakers.

### Financial clusters — stages and policy considerations
- Typical progression:
  - Trade links precede financial interconnections.
  - Early financial linkages: foreign bank penetration causing one-way spillovers.
  - Later: two-way spillovers via international equity and debt markets as legal/governance standards mature.
- Policy trade-offs:
  - Cross-border bank expansion from gatekeeper to host can constrain host/home policies; sharing supervisory information, coordinating prudential standards, and cooperating on supervision/resolution can maximize benefits.
  - Absent coordination, hosts may enact defensive policies; regulatory arbitrage can skew benefits toward firms and costs toward taxpayers.
  - Reducing clustering caused by regulatory arbitrage could be a net gain.

---

### Surveillance implications and recommended actions
- Surveillance challenge: incorporate interconnectedness into Fund surveillance to better assess risk and policy implications.
- Policy-oriented measures and actions:
  - Context setting: account for partner, gatekeeper, and core policies in Article IV and FSAP discussions.
  - Surveillance of clusters: consider broader cluster surveillance (pilot basis), including thematic reports to inform Article IVs, and include in multilateral/regional products where appropriate.
  - Cooperation: establish modalities for information sharing (e.g., financial supervision) and policy dialogue/cooperation within clusters, both in crisis and normal times.
  - Circuit breakers: identify “pressure points” and “circuit breakers” (e.g., Italy) to focus surveillance and buffer-building to limit shock propagation.
  - Data: exercise is data dependent; mapping financial interlinkages across G-SIFIs, markets, sectors, and countries is constrained by information gaps; efforts needed to close gaps (G20 Data Gaps Initiative).
- Further work:
  - Pilot studies to assess value added for surveillance and vulnerability exercises.
  - Explore alternative approaches (structural models) and parallel workstreams to better understand relationships among interconnectedness, stability, and policies.

*Content from IMF chapter "ENHANCING SURVEILLANCE: INTERCONNECTEDNESS AND CLUSTERS" (selected sections 1, 17, 27–46) from the PDF content provided.*

### 1. Structural Features of Global Trade and Financial Flows Network: Gatekeepers _______________  15

### 1. Structural Features of Global Trade and Financial Flows Network: Gatekeepers

### Motivation
- Interconnectedness matters: rising cross-border trade and finance have delivered benefits (access to new markets, cheaper funding, more efficient allocation of capital) but also pose risks—shocks in one part of the system can be transmitted widely and quickly.
- Fragility and limited policy buffers are expected to persist in the advanced economy core of the global financial system, leaving potential for systemic volatility high.
- The Fund’s Triennial Surveillance Review (IMF, 2011c) called for enhancing understanding of interconnectedness and incorporating this into risk and policy analysis.
- Understanding remains limited: no satisfactory comprehensive theory backed by detailed empirics on relationships among interconnectedness, stability, and policies; data gaps on cross-border financial exposures and G-SIFIs complicate analysis.
- Need for a framework that tracks direct, indirect, and common linkages across trade, banking and portfolio exposures, FDI, policies, payments systems, and common characteristics.

### This paper — two-part approach
- Architecture (Stage I): use cluster (network) analysis (clique percolation) to map the architecture of cross-border trade and financial interconnectedness, revealing:
  - a global core (mainly, the systemic-5),
  - clusters of economies with closer internal connections (e.g., Nordic-Baltics),
  - gatekeepers/connectors that link clusters to one another and the core to clusters (e.g., China appears as a gatekeeper to the Asian supply chain).
- Applications (Stage II): select case studies to analyze how cluster functional/economic features shape shock propagation and policy implications; context matters (no general relationship between rising interconnectedness and output synchronicity).

### Key hypotheses and implications generated
- Gatekeepers can act as dampeners or amplifiers of shocks; sensitivity analyses identify gatekeepers as possible pressure points or “circuit breakers”.
- Situating an economy within its cluster(s) can:
  - facilitate sharper assessment of shock propagation and risks;
  - highlight which partners and which policies/buffers may matter for systemic stability;
  - provide a basis for dialogue on macroeconomic, macro-prudential, and capital flow management policies within clusters.
- Caveat: cluster analysis generates hypotheses that require confirmation through deeper economic analysis to avoid type I and type II errors.

### Framework — Two-stage analytical approach
- Stage I: Mapping the Network Architecture (behavior- or risk-neutral)
  - Problem: dense bilateral trade and financial connections produce a high-dimensional, visually opaque network (Figure 1: Dense Interconnected Network).
  - Method: clique percolation clustering condenses bilateral trade and financial data into overlapping subgroups/clusters of economies; overlapping clusters preserve multi-affiliation (e.g., China with Asian supply chain and with advanced economies).
  - Stylized architecture (Figure 2): overlapping clusters, a core (intersection of many clusters), and gatekeepers linking clusters and core.
  - Generated hypotheses about roles:
    - Core: shocks and policies in the core can have global spillovers (see IMF, 2011b; IMF, 2011e).
    - Clusters: shocks may have larger initial impact within a cluster given denser connections (Europe highlighted as a pertinent example).
    - Gatekeepers: may transmit or dampen shocks; example—Austrian banks borrowed wholesale and passed through low funding costs to some CEE economies (hypothesized as conduits); Austria’s fiscal recapitalization appears to have limited deleveraging and acted as a dampener.

- Stage II: Analyzing Clusters and Evaluating Hypotheses
  - Need to delve into functional characteristics and economic context of clusters to test Stage I hypotheses.
  - Interconnections can:
    - dissipate certain shocks (e.g., easier substitution),
    - amplify or propagate others (e.g., via gatekeepers).
  - For enduring clusters, shocks may be perceived to be dissipated by gatekeepers; otherwise cluster membership may change as links diminish.
  - Existing literature is ambiguous: effects of trade and financial integration on output synchronization or volatility depend on:
    - extent and nature of integration (inter- vs. intra-industry trade),
    - types of shocks (industry-level vs. common shocks),
    - levels of economic and financial development (ability to diversify risk),
    - type of financial clustering, and prudential frameworks (Background Paper, Chapters I and VI).
  - Initial considerations for cluster analysis (Background Paper, Chapter VIII) should inform further structural/event studies.

### Policy considerations and potential actions
- Trade-related implications:
  - When close trade linkages produce synchronized business cycles, there may be less loss from foregoing monetary policy independence (OCA logic), strengthening the case for policy coordination.
  - Where cycles are not synchronized, policy flexibility is desirable to avoid hindering adjustment and creating tensions.
  - Surveillance can highlight policy “tensions” and levers by revealing cluster-level cycle relationships (Background Paper, Chapter II provides stylized facts on output synchronization).
- Finance-related implications:
  - Close financial integration can drive capital inflows and expansion in upturns and sudden stops or reversals in downturns.
  - Information sharing and coordination of policies to manage inflows may be warranted:
    - reactions to inflows in one country can divert flows elsewhere,
    - coordination with source countries of inflows can enhance awareness of external effects and improve policy effectiveness (IMF, 2011e).

### Caveats and limits
- Results are sensitive to data quality and quantity; more comprehensive data (notably on financial exposures) would sharpen findings.
- Cluster analysis is a step toward understanding interconnectedness but is not definitive; hypotheses must be tested with detailed economic analysis and models to infer causal roles of core economies and gatekeepers.

*Source: IMF chapter "1. Structural Features of Global Trade and Financial Flows Network: Gatekeepers" from the PDF content provided.*

### 17.      Data. For the cluster analysis, detailed bilateral trade data are used from the IMF’s Direction

### _031512 - 17.      Data. For the cluster analysis, detailed bilateral trade data are used from the IMF’s Direction

### Data and methodology
- Data sources:
  - IMF’s Direction of Trade Statistics on aggregate data.
  - UN Comtrade on various commodities, capital goods, and manufacturing goods.
  - BIS consolidated and locational banking statistics.
  - IMF’s Coordinated Portfolio Investment Survey (CPIS).
  - IMF’s Coordinated Direct Investment Survey (CDIS).
- Data vintage: all data are as of end-2009.
- Inclusion threshold: only bilateral connections above 15 percent of a country’s overall connections in that category are used (threshold determined endogenously to maximize overlaps across clusters).
- Example: country A’s trade in capital goods with country B is included only if that trade is above 15 percent of A’s overall trade in capital goods.
- Aggregation: bilateral links of each economy are summed across trade and financial dimensions, with links weighted by their relative shares in respective datasets to avoid adding absolute numbers across differently scaled datasets.
- Gatekeeper identification: iterative process—if removing an economy causes a cluster to drop out, that economy is identified as connecting that cluster to the network.

### Finance and trade architectures (findings)
- Finance: nested core-periphery structure; handful of economies intermediate global finance (core connected to all other economies and present in almost all clusters).
  - Example dataset illustrated: BIS locational banking statistics (Figure 3A).
- Trade: core of some of the largest economies but in a non-nested structure with fewer overlaps among clusters (Figure 3B).
- Heterogeneity by product: architectures vary by type of good (e.g., machinery vs. commodity), producing different numbers/types of clusters and overlaps.
- Implication: varied roles of economies across products/networks point to different potentials to originate, amplify, or propagate sectoral or aggregate shocks.

### Overall aggregated architecture (findings)
- Key points from aggregated trade and financial network (projected onto 2-D in Figure 4):
  - there are a large number of overlapping clusters;
  - the global core economies, comprising mainly the systemic-5, are connected to most other economies and many clusters;
  - in several clusters, there are more than one gatekeeper; multiple gatekeepers point to potential negative feedback effects within and between clusters as shocks must pass multiple “gates” to spread;
  - economies that do not appear in any individual clusters are nevertheless part of the entire global system (may reflect limited data on some dimensions, such as cross-border finance).

### Clusters and gatekeepers (findings and illustrative examples)
- Core economies identified by density of connections, centrality, and presence in several clusters.
- Intuitive clusters and gatekeepers identified, with some illustrative examples:
  - Asian cluster around China and the United States.
  - Euro Area cluster around Germany.
  - Austria linking CEE economies and South Eastern Europe (Bosnia and Herzegovina, Croatia, and Slovenia) to other large European economies (CESEE).
  - Sweden linking Latvia and Estonia to other large European economies.
  - Central America clustered around the United States.
- Surprises in cluster assignments noted (e.g., Lithuania closer to Poland than to Estonia and Latvia; Algeria, Egypt, Lebanon, and Tunisia closer to European clusters).
- Table 1 results (aggregate network):
  - total groups of gatekeepers: 21;
  - groups that cut across regions: 18;
  - groups that fit regional categories: 3 (two connect advanced Europe with emerging Europe; one relates to Southern Africa).
  - United States identified as a gatekeeper to 16 groups.
  - China identified as a gatekeeper to 12 groups.
- Selected group-level implications:
  - Group P: problems in Greece can have significant spillover effects beyond its cluster because gatekeepers—Italy and Germany—are adversely impacted; Italy’s stress could spread across Europe, the Middle East, Central Asia, and sub Saharan Africa.
  - Group O: Italy is a gatekeeper (with Germany and Austria) into some CESEE economies; stresses in Italy could have direct adverse impacts.
  - Group N: Egypt and some Middle Eastern countries could be adversely impacted through strong, direct linkages to Italy.
  - Group J: Italy as conduit to Belgium, France, Germany, and the Netherlands—secondary reverberation effects among these core European countries could further impact other clusters such as the CESEE; robust domestic policy buffers could contain broader spillovers.
  - Group Q: CIS economies may not be as susceptible to secondary reverberation effects except via certain other emerging market economies.
  - Group U: Italy, with large advanced and emerging economies, links a number of sub Saharan African economies to the global network; attenuation of impacts depends on policy space of other gatekeepers and on potential adverse feedback loops.
  - Group K: Hong Kong SAR may serve as an equity financing center intermediating Asian economies and U.S. investment via G-SIFIs.

### Durability over time
- Cluster analysis conducted over 6 different time periods from 2000 to 2009.
- Broad stability of resulting cluster structure, with exception:
  - China grew rapidly and became more central to many clusters.
  - Consequence: Asian supply chain became both more central to global trade and more concentrated.
- Caveat: data limitations (particularly on financial exposures) dominate; results may need revisiting as more complete data become available.

### Other properties and quantitative indicators
- Output correlations:
  - average real GDP growth correlation over the past 20 years is 0.37 within clusters versus 0.12 for the world.
  - Interpretation: within-cluster co-movement is three times higher than world average but not more than a standard deviation apart; clusters may often mitigate rather than amplify local shocks.
- Context dependence:
  - amplification or dissipation of shocks depends on nature of shock, economic/financial depth of cluster members, and policies.
  - Financial shocks originating in or impacting the core (nested in nearly all clusters) may have widespread effects.
  - Real/trade shocks may have more localized effects due to fewer overlaps in trade clusters.
  - Real sector and financial diversification and capacity for counter-cyclical policy matter for transmission.

### Sectoral and granular cluster analysis (recommendations / extensions)
- Aggregated network can be decomposed into sub-networks (e.g., machinery, capital goods, particular financial exposures such as through G-SIFIs).
- For countries concentrated in specific sectors/assets (e.g., commodity exporters), analyzing granular clusters and gatekeepers can yield more detailed information.
- Augmentations suggested:
  - include membership in formal currency areas;
  - construct clusters based on correlated asset prices to capture similarities (rather than direct exposures) and better address contagion.

### Vulnerability / stress testing application
- Sensitivity exercise: impose exogenous/mechanical reductions of links and compare resulting clusters/gatekeepers with pre-perturbed system.
- Endogenous minimum threshold for links is determined by methodology; links pruned below this threshold leave cluster features unchanged.
- Illustrative stress applied: uniform stress to all trade and financial links of 15 percentage points causes some links to fall below the threshold (Figure 7).
- Interpretation of results (as described for Figure 7):
  - each country in figure associated with two numbers: left = share of envisioned shock transmitted to it (incoming, inward directional links in proportion to total); right = share of its effects on others (outward directional links in proportion to total).
  - Italy identified as a “pressure point” (envisaged shock to Italy significantly affects multiple regions; e.g., Egypt’s and Greece’s links fall below the threshold). If Italy has adequate buffers it can act as a “circuit breaker”.
  - Some partners are recipients of shocks directly and indirectly; others, including core countries, continue to transmit shocks, potentially dampening or amplifying them (examples: United States, Germany, France, Belgium; also Gabon and Cameroon within respective clusters).
  - United Kingdom appears as a shock absorber from Europe and as a conduit for shocks to the United States because of its role as global intermediator of U.S. dollar assets; real effects manifest in economies that issued or used those funding vehicles rather than in the United Kingdom itself.
  - China’s density lies more within Asia—stickiness in trade relationships could mute impact of certain shocks; however, stress in China could produce material effects on its clusters, including the United States.
- Caveat: the mechanical exercise maps potential transmission routes and feedback loops; it is not a quantitative gauge of relative intensity of shocks or of actual propagation—careful identification of shocks and robust empirical work are needed.

*ENHANCING SURVEILLANCE: INTERCONNECTEDNESS AND CLUSTERS, INTERNATIONAL MONETARY FUND*

### 27.      Structure. The Asian cluster is a vertically integrated supply chain that comprises: China,

### 27.      Structure. The Asian cluster is a vertically integrated supply chain that comprises: China,

### Structure
- Composition: China, Hong Kong SAR, Indonesia, Japan, Korea, Malaysia, Philippines, Singapore, Thailand, and Vietnam (Background Paper, Chapter V).
- Role of gatekeepers:
  - China identified as a gatekeeper given its economic size, growing trade intensity with the region, and centrality in global trade (Table 1, group H).
  - Japan previously played an important gatekeeper role and remains a key player in Asian trade.
  - Hong Kong SAR and Singapore link the cluster to global financial centers (United States and United Kingdom) and intermediate financial flows from a wider set of countries to the rest of the region (Table 1, group K).
- Dominant linkage: Trade linkages appear to dominate within the cluster.

### Cycles
- Linkage to global electronics cycle:
  - Cluster is closely linked to the global electronics business cycle due to vertically integrated production (especially machinery and electronics) and relatively high export similarity across several countries.
- China’s evolving role:
  - China is increasingly an important source of demand in the region, as illustrated by its expansionary policies in 2008–09; it is no longer merely a processing center (IMF, 2011b).
- Synchronization:
  - Business cycles of countries in the cluster are quite closely—though not entirely—synchronized with China’s.

### Policies
- Exchange rate co-movement:
  - Considerable co-movement of nominal exchange rates in the region, attributed to vertical integration, high export similarity, and co-moving business cycles.
- Policy implications for downstream countries:
  - For countries downstream in the supply chain (e.g., Malaysia, Thailand, the Philippines, Vietnam), preserving their place in the production structure relative to China could entail maintaining relative currency stability vis-à-vis each other.
  - Consequence: Availability of the exchange rate instrument as a policy tool could become limited when facing idiosyncratic shocks or divergent cycles, shifting adjustment to other policy instruments.

### Example: China’s July 2005 exchange rate reform
- Event:
  - In July 2005, Chinese authorities abandoned the peg to the dollar and announced reforms to allow greater flexibility.
- Outcome by end-2008:
  - Despite sustained capital inflows and large foreign exchange reserve accumulation, the renminbi appreciated by a cumulative 20¼ percent vis-à-vis the U.S. dollar by end-2008.
- Policy context:
  - China’s investment-driven growth increased absorptive capacity while inflationary pressures remained relatively contained, providing space for gradual appreciation as monetary policy relied heavily on administrative measures.

### Policy responses in regional countries (pre-crisis period, amid surge in capital inflows starting 2003–04)
- Malaysia:
  - Abandoned its peg to the U.S. dollar coinciding with China’s announcement.
  - Exchange rate remained managed and movements closely mirrored renminbi appreciation until mid-2008.
  - Fiscal consolidation accommodated a slower pace of currency appreciation despite surging capital inflows.
- Thailand:
  - Despite relatively prudent macroeconomic policies and renminbi appreciation, Thai currency appreciation outpaced other regional currencies against the dollar and the renminbi.
  - Resulted in imposition of capital controls in end-2006; some measures reversed almost immediately and others removed gradually over the next 1½ years.
  - Prudential measures became an additional instrument where traditional instruments were constrained by cluster policies.
- Philippines:
  - Fiscal dominance constrained monetary policy’s ability to mitigate surge in capital inflows.
  - Authorities opted for greater nominal exchange rate appreciation to contain inflationary pressures.
  - Result: Philippine currency appreciated the most against both the dollar and the renminbi by 2008.

### Cluster-based approach
- Policy space augmentation:
  - Use of macro-prudential and capital flow management policies could be invoked to ‘augment’ policy space when exchange rate flexibility is limited.
- Surveillance and cooperation:
  - A cluster-based approach to surveillance can clarify policy trade-offs and tensions, and indicate whether policy cooperation within the cluster—such as sharing information on the external effects of policies and possible coordination—could mitigate these trade-offs.

---

### Saudi Arabia—Commodity Cluster

### Structure and context
- Evolving trade links:
  - Commodity exporters have seen increased demand driven, at the margin, by EMs such as China; trade links of oil producers like Saudi Arabia have evolved (Table 1, group G).
  - Saudi Arabia’s gatekeepers now include the United States, China, and Korea, besides Japan.
- Financial sector development:
  - Domestic financial deepening in Saudi Arabia strengthened monetary transmission.
  - Given the U.S. dollar peg, U.S. monetary conditions have increased influence on the Saudi non-oil sector.
  - Result: Growth and fiscal revenues influenced by EMs, while monetary policy conditions continue to be guided by U.S. conditions.

### Event (September 2007–October 2008)
- U.S. monetary policy:
  - U.S. Federal Reserve lowered policy rate from 5.25 percent to 1 percent.
- Oil prices:
  - Crude oil prices surged from over $65 per barrel in mid-2007 to over $130 per barrel by summer 2008, driven partly by increased EM demand.
- Saudi specifics:
  - Rise in oil prices (2004–2007) increased government oil revenues, spending, growth, and inflationary pressure.
  - To prevent speculative capital inflows and maintain the dollar peg, the Saudi Arabian Monetary Agency cut its policy rate from 5 percent in October 2007 to 2 percent by mid-2008.
  - Annual credit growth surged from 6 percent in early 2007 to over 30 percent in July 2008.
  - Higher global commodity prices and a depreciating dollar contributed to inflationary pressure, resulting in double digit inflation by mid-2008.

### Cluster implications
- Trade-offs:
  - If global oil prices move counter-cyclically with U.S. monetary policy, sharp trade-offs emerge between maintaining the dollar peg and managing domestic credit growth and demand.
- Policy options:
  - Importance of strong fiscal management.
  - Further refinement of macro-prudential instruments to influence monetary conditions independently of interest rate policy.

---

### Financial Clusters

### Stage of development (Table 2)
- Typical progression:
  - Early stage: Cross-border real sector interconnections (trade) precede financial interconnections.
  - Early financial linkages: Penetration of domestic credit markets by foreign banks causing one-way spillovers from foreign banks’ home countries.
  - Later stage: Government and private sector activity in international equity and debt markets follows when legal, corporate governance, and disclosure standards meet investor due diligence needs; linkages can evolve from one-way to two-way spillovers.
  - Advanced economies: Connected via cross-border bank ownership, bank funding flows, and market borrowing/lending in international capital markets.

### Exploiting synergies or arbitrage?
- Drivers:
  - Some clusters arise from institutions exploiting economies of scale and scope; others from opportunities for tax and regulatory arbitrage (Background Paper, Chapter VI).
- Policy considerations:
  - Where banks expand from gatekeepers (“home”) to hosts to exploit diversification or serve corporate clients, spillovers may constrain host and home country policies.
  - To maximize benefits, authorities could share supervisory information, coordinate prudential standards, and cooperate on supervision and resolution.
  - Absent coordination, hosts may implement defensive policies limiting cluster participation.
  - Policy initiatives that encourage cross-country financial integration without harmonized prudential standards and cross-border supervisory cooperation could enable regulatory arbitrage, skewing benefits toward firms and costs toward taxpayers.
  - Reducing clustering generated by closure of regulatory arbitrage opportunities could represent a net gain to all countries in such cases.

---

### Nordic-Baltic Countries—A Financial Cluster

### Structure and context
- Composition:
  - Gatekeepers: home countries of Nordic banks (Denmark, Norway, Sweden).
  - Hosts: Baltic countries (Estonia, Latvia, Lithuania).
  - Gatekeepers connected to the global core via large funding operations in wholesale capital markets (Background Paper, Chapters VI and VII).
- Pre-crisis dynamics:
  - Parent banks passed global liquidity to hosts, fueling economic and credit/asset booms.
  - Host markets became vital to group liquidity/solvency; supply and cost of funds driven by global financial centers rather than cluster policies.
  - In the EU context of open capital accounts, this clogged the credit channel of monetary policy.
  - Hindsight: stronger countercyclical stance in the Baltics could have arrested overheating; Nordic financial stability policies failed to address rising wholesale funding and intra-group exposures.

### Crisis shock and response
- Funding withdrawal:
  - Withdrawal of funding by Swedish banks amplified the global shock; transforming short-term wholesale funding into longer-term Baltic loans became unviable.
- Reliance on Sweden:
  - Limited countercyclical scope in Baltics, solvency risk for systemic Swedish banks (SEB, Swedbank), and refinancing pressure post-Lehman meant reliance on Sweden’s policy space to back-stop Baltic exposures.
- Swedish actions:
  - Swedish authorities provided liquidity in kroner and foreign currencies via swap arrangements, directly supported pressured banks, and participated in the Vienna Initiative.
  - Latvia had an IMF-supported program.
  - These measures stabilized markets, eased deleveraging, and prevented fire-sale dynamics; otherwise, Baltic austerity and the cost of bailouts would have been larger.

### Gatekeeper role analysis
- Structural VAR evidence:
  - Comparing unrestricted impulse responses with responses where the gatekeeper’s money market rate is held constant shows Swedish monetary policy generally lowered spillovers from negative shocks to money market rates and GDP growth in the Baltics, especially Estonia and Latvia (Background Paper, Chapter VII).
  - Conditional responses vary by strength of financial linkages and policies in destination countries; use of fiscal buffers in Estonia could have reduced measured impact of accommodative Swedish policy.

### Cluster policy lessons
- Policy menu to better serve macro-financial stability:
  - Strengthen fiscal stabilizers or adopt structural measures in Baltic countries to contain credit booms and limit Nordic banks’ exposures, reducing contingent liabilities for Swedish taxpayers.
  - Nordic home authorities should monitor and curb growth in intra-group exposures; this requires information sharing and coordination.
  - Gatekeeper actions can both amplify transmission initially and, via policy buffers (fiscal space), later act as a circuit-breaker mitigating crisis impact.

---

### V. IMPLICATIONS — Summary
- Two-stage approach:
  - Stage 1: Network analysis mapped cross-border trade and finance architecture, revealing a global “core” and distinct “clusters” with strong intra-cluster connections; clusters connect via “gatekeepers” that can amplify or dampen shocks and affect policies within the cluster.
  - Stage 2: Case studies highlighted cluster-specific issues:
    - Commodity exporters: cluster perspective essential for analyzing trade-offs when rising demand comes from dynamic EMs while policy linkages remain with slower-growing advanced economies.
    - Vertically integrated supply chains (Asian trade cluster): cluster view clarifies when exchange rate flexibility and other instruments may be limited for some economies.
    - Financial clusters: sharing information across the cluster could improve risk understanding and policy effectiveness.
- Gatekeepers:
  - Can amplify or dampen external shocks; structural and behavioral relationships warrant further exploration, possibly via pilot studies in surveillance of countries and clusters, to draw firmer inferences and policy implications.

*Italic line: Content from IMF chapter "ENHANCING SURVEILLANCE: INTERCONNECTEDNESS AND CLUSTERS" (selected sections 27–45).*

### 46.      Surveillance. The 2011 Triennial Surveillance Review called for progress in incorporating the

### 46. Surveillance

### Surveillance challenge and objectives
- The 2011 Triennial Surveillance Review called for progress in incorporating the fact and implications of interconnectedness into the Fund’s surveillance, with an increased focus on risk assessments and a broader sweep of economic and financial stability (IMF, 2011c).
- Currently, little is done systematically to situate the analysis of an economy’s policies in the interconnected trading and financial system, or to draw implications for overall stability.
- The surveillance challenge is to:
  - understand the architecture of interconnectedness;
  - characterize its functional aspects and implications for shock propagation and policies; and
  - track the evolving nature of the connections.
- This paper takes a first step toward understanding the architecture and implications of interconnectedness. Once the architecture and the functional and behavioral underpinnings are better identified and assessed, implications for surveillance could be drawn.

### Implications for surveillance (policy-oriented measures and actions)
- Context setting.
  - Taking account of the policies of partners—in a cluster, of the gatekeeper, and in the core—has the potential to nuance the understanding of shock propagation, policy space, and stability.
  - Providing such a context in the Article IV and Financial Sector Assessment Program (FSAP) discussions of countries is a first step.
  - Grouping the Article IV and, where relevant, the FSAP discussions by cluster at the Executive Board is a related step.
- Surveillance of clusters.
  - Broader surveillance of clusters could be considered, if first on a pilot basis, beyond the formal currency unions, such as the Euro Area.
  - Interconnectedness-based clusters such as the Asian supply chain or Nordic-Baltic cluster can give rise to stability and policy challenges at the country level that are better assessed at the cluster level.
  - Such surveillance could be conducted through thematic reports as background to the relevant Article IVs.
  - They could also be covered, where appropriate, in the multilateral surveillance and regional products.
- Cooperation.
  - In time, and if value is gleaned from this approach, modalities for cooperation may also need to be established within the cluster, if these are not already in place.
  - Information sharing (e.g., on financial supervision) would be essential.
  - Policy dialogue and cooperation may also become necessary, not only in crisis time (e.g., the Vienna Initiative) but also in normal times.
- Circuit breakers.
  - Identifying potential “pressure points” and “circuit breakers” (such as Italy) would point to the need to focus particular attention—in terms of surveillance and building buffers/the safety net—on them so as to enhance the discussion of systemic stability.
  - Ensuring the robustness of these circuit breakers could limit the extent to which adverse shocks spread through the system.
- Data.
  - This is a data dependent exercise.
  - The ability to map financial interlinkages across G-SIFIs, markets, economic sectors, and countries, while essential to the identification of clusters and ongoing assessments of developments therein, is severely constrained by information gaps.
  - These gaps correspond to the unavailability of relevant information—on a comprehensive and consistent basis across the membership—in the public domain.
  - Efforts would continue to be needed to close these gaps, including through the G20 Data Gaps Initiative.

### Further work
- Further work is needed on the relationships among interconnectedness, stability, and policies. This work could proceed on multiple, parallel tracks.
- Pilot approach.
  - The above approach could be piloted in a few cases, to discern and sharpen the value added that could be brought to bear in surveillance of better understanding how interconnectedness affects risk assessments and policies.
  - To the extent that these relations are context dependent, knowledge of the architecture—of the core, clusters, and gatekeepers, which evolves over time—can focus attention on the circumstances in which shocks can be dampened or amplified and propagated.
  - The above work could also be deployed in the vulnerability exercises.
- Alternative approaches.
  - Alternative means of understanding shock propagation could be explored, including through structural models.

*Source: _031512 - 46.      Surveillance.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2012/_031512.pdf_