## _wp1657

## Source details

**Canonical URL:** [_wp1657](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1657.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1657.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1657.pdf.json)

---

### I. Introduction and motivation
- Focus: quantify how shocks from Asian equity markets contribute to other countries’ equity returns and volatilities over time, and assess which markets are major sources and recipients of shocks.
- Context and examples:
  - Financial integration within Asia remains relatively low: "only about 20 percent to 30 percent of cross-border portfolio investment and bank claims are intraregional."
  - Trade intraregional share: "53.3 percent of total Asia’s merchandise flows end up in Asia"; Euro area intra-regional trade share "68.8%".
  - Example shock episode: China’s stock market fell "8.5 percent" on August 24, 2015; A-shares in the Shanghai Stock Exchange lost "one third" of their market value.
- Policy relevance: informs regional liquidity support arrangements (Chiang Mai Initiative Multilateralization) and ASEAN capital market integration guidelines; useful for crisis monitoring.

### II. Data and empirical setup
- Data coverage:
  - Daily nominal local-currency stock market indexes from "January 1996 to October 2015" (Haver Analytics).
  - Sample: 13 Asian economies (Australia, China, Hong Kong SAR, India, Indonesia, Japan, Korea, Malaysia, New Zealand, the Philippines, Singapore, Taiwan Province of China, Thailand), four advanced countries (United States, United Kingdom, Germany, France) and three emerging markets from other regions (Brazil, Mexico, Turkey). Additional estimations included Europe and South Africa (sample from "June 24, 2002").
- Return and volatility construction:
  - Asset returns: change in the log-price, Friday-to-Friday (nominal return).
  - Volatility: Parkinson (1980) estimator using daily high and low prices; annualized daily volatility reported as percent standard deviation.
- Descriptive statistics: weekly returns (Table 1) and daily volatility (Table 2) reported in the source.

### III. Measures and methodology
- Framework: Diebold and Yilmaz connectedness measures (Diebold and Yilmaz 2009, 2012, 2014) using forecast-error variance decompositions from VARs.
- Identification:
  - Generalized VAR (Koop, Pesaran, and Potter 1996; Pesaran and Shin 1998) used to obtain variance decompositions invariant to variable ordering and allowing correlated shocks.
- Key definitions:
  - Total connectedness: sum of cross-variance shares (fractions of H-step-ahead forecast-error variances of variable i due to shocks to variable j, j ≠ i).
  - Directional connectedness:
    - "From" connectedness: shocks received by vector i from all other vectors j.
    - "To" connectedness: shocks transmitted by vector i to all others.
  - Net connectedness: difference between "To" and "From" connectedness for each market.
  - Dynamic connectedness: time-varying connectedness computed using rolling-window VAR estimates.
- Baseline estimation choices:
  - VARs with "four lags".
  - Connectedness calculated for a "10-week horizon".
  - Rolling-window baseline: "200-week window".
  - Robustness alternatives: different lags ("eight months"), forecasting horizons ("15, 20 weeks"), and windows ("100 weeks").

### IV. Main empirical findings
- Aggregate trends:
  - "There has been a substantial increase in the aggregate equity return and volatility connectedness indexes since the global financial crisis."
  - Post-crisis: a mild decrease in recent years after a sharp rise in the aftermath of the global financial crisis.
- Returns versus volatilities:
  - Equity returns connectedness: smoother upward trending behavior.
  - Equity volatilities connectedness: displays spikes (burst behavior) in addition to the trend; both measures peaked during the global financial crisis.
- AEs versus EMs:
  - Since the global financial crisis:
    - Advanced economies have increasingly become net "receiver" of shocks.
    - Emerging market economies, particularly China, ASEAN-5 (Indonesia, Malaysia, the Philippines, Singapore, and Thailand), and Korea, have increasingly become net sources of shocks.
  - Drivers:
    - AEs: decline in "contribution to others" and increase in "contributions from other countries".
    - EMs: increase in "contribution to other economies".

### V. Role of China and recent dynamics
- China’s evolving role:
  - China’s net contribution to other markets rose, especially after heightened volatility since summer 2015; became a growing net source (transmitter) of shocks to economies inside and outside the region, including Hong Kong SAR, U.S., and European markets.
  - Japan’s contribution to other equity market movements declined over the same period.
- Recent episode details:
  - Net connectedness index (CI) from China to other equity markets increased sharply since the second half of 2015.
  - The steep increase in CI since mid-2015 is more pronounced than during the "taper tantrum" (May-July 2013), suggesting "contagion" effects rather than only direct financial linkages.

### VI. China–Hong Kong SAR nexus and sectoral heterogeneity
- Hong Kong SAR linkage:
  - Hong Kong SAR is a regional financial hub; ties to China increased as China liberalized its financial markets and opened its financial account.
  - Initiatives (e.g., Connect scheme) likely boosted financial integration and created new transmission channels.
- Methodology for Hong Kong SAR analysis:
  - VARs estimated with daily data and "four lags".
  - Spillovers for a "10-week horizon".
  - Rolling-window estimations; sample period June 2002–August 2015 for country analysis; sectoral indexes sample Feb 2005–Aug 2015.
  - Indexes: Dow Jones Shanghai (China), Hang Seng Index (Hong Kong SAR), Hang Seng Composite Industry Index (Hong Kong SAR).
  - Baseline: "200-week window"; robustness to "eight months" lags, "15, 20 weeks" horizons and "100 weeks" window.
- Results:
  - Interlinkages between mainland China and Hong Kong SAR equity markets are strong and growing.
  - China’s equity markets explain a larger share of Hong Kong SAR’s equity returns and volatility than other economies.
  - Singapore’s average connectedness with the Mainland similar to Malaysia and Korea.
  - Return CI rose with a spike in late 2014; volatility CI peaked in 2011.
  - Net spillovers from China to Hong Kong SAR equity volatility have been negative at times, implying Hong Kong SAR shocks also hit China.

- Sectoral heterogeneity (China → Hong Kong SAR):
  - Sectors with sharper rises in CI: Financials; Properties and Construction; Consumer goods.
  - Information and Technology sector experienced a noticeable, albeit less drastic, increase.
  - Financials and Properties and Construction have decoupled from other sectors, first from Information and Technology, and more recently from Consumer goods — possibly reflecting stronger real linkages and banks’ exposure to China’s property markets.
  - CI magnitude for sector indexes reaches up to at least 16 in the reported timeseries.

### VII. Quantitative connectedness summary (selected totals from Appendix)
- Equity Returns Connectedness (Time Period: 06/30/2002-11/01/2015) — selected entries:
  - "To others" totals: USA To others 110%; GBR 118%; EUR 99%; CHN 84%; HKG 12%; KOR 108%; TWN 92%; SGP 64%; IND 84%; ZAF 49%.
  - Total Connectedness (including own): USA 124%; GBR 130%; EUR 114%; CHN 100%; HKG 62%.
  - Example "From Others" and "Net": USA From Others 78% Net 24%; GBR From Others 68% Net 30%; CHN From Others 58% Net 0%; HKG From Others 50% Net -38%.
- Equity Volatilities Connectedness (Time Period: 06/30/2002-11/01/2015) — selected entries:
  - "To others" totals: USA 111%; GBR 133%; EUR 79%; CHN 49%; HKG 77%; SGP 130%; IND 102%; ZAF 67%.
  - Total Connectedness (including own): USA 131%; GBR 150%; EUR 99%; CHN 68%; HKG 125%.
  - Example "From Others" and "Net": USA From Others 68% Net 31%; GBR From Others 83% Net 50%; CHN From Others 48% Net -32%; HKG From Others 15% Net 25%; ZAF From Others 88% Net 74%.

### VIII. Appendix dynamic ranges (selected axis values and observed ranges)
- Appendix Figure 5 — Contributions to Others (Equity Volatility) examples:
  - USA: 80, 85, 90, 95
  - CHN: 20, 40, 60, 80
  - HKG: 70, 75, 80, 85, 90, 95
  - THA: 0, 20, 40, 60, 80, 100
- Appendix Figure 6 — Contributions from Others (Equity Returns) examples:
  - USA: 75, 80, 85, 90, 95
  - CHN: 40, 50, 60, 70, 80
  - THA: 65, 70, 75, 80, 85, 90
- Appendix Figure 7 — Contributions from Others (Equity Volatility) examples:
  - CHN: 50, 60, 70, 80, 90, 100
  - IDN: 20, 40, 60, 80, 100
  - MEX: 60, 70, 80, 90, 100
- Appendix Figure 8 — Net Connectedness (Equity Returns) axis examples:
  - CHN: -30, -20, -10, 0, 10
  - JPN: -20, -15, -10, -5, 0, 5
  - NZL: 0, 5, 10, 15, 20, 25, 30
- Appendix Figure 9 — Net Connectedness (Equity Volatility) axis examples:
  - CHN: -40, -30, -20, -10, 0, 10
  - THA: -60, -40, -20, 0, 20
  - MYS: 0, 10, 20, 30, 40

### IX. Robustness and sensitivity
- VAR specification robustness:
  - Results robust to different windows and lags; main results derived from four lags, 10-week horizon, and 100-week rolling window estimations; baseline uses 200-week window for dynamic CI.
  - Appendix Figure 10 compares H=10 weeks (baseline) and H=20 weeks: Total Connectedness ranges shown for both H=10 weeks and H=20 weeks — 70, 72, 74, 76, 78, 80, 82, 84, 86, 88, 90.
- Exogenous global factors:
  - Tested exogenous variables: VIX; Citigroup Economic Surprise Index; Global commodity prices (S&P GSCI Spot Index); Bloomberg commodity spot index; monthly Global PMI (JP Morgan); monthly Global PMI Manufacturing only output (JP Morgan).
  - Exogenous factors entered separately and jointly.
  - Finding: results change little when exogenous global factors are included; CI can be interpreted as country-originated shocks netting out common global-factor shocks.

### X. Policy implications and concluding remarks
- Key conclusions:
  - Asian emerging markets, particularly China, have become important sources of financial shocks with increasingly large spillovers to other economies, including advanced market economies.
  - China accounts for about one-third of global growth and is a major player in global trade and, increasingly, global finance.
  - Since the global financial crisis, China has been increasingly a net "giver" of financial shocks to the Asia-Pacific region.
  - Ongoing liberalization of China’s domestic financial markets and capital controls, and RMB internationalization, imply China’s financial-channel importance will continue to rise.
  - Intraregional connectedness among emerging markets has increased and is expected to grow with strengthening trade and financial linkages.
- Policy recommendations:
  - Use time-varying connectedness metrics (total, directional, net) for monitoring spillovers and crisis detection.
  - Financial regulation and supervision should emphasize regional interlinkages and identify regulatory gaps across markets and borders that could enable shock transmission.
  - Consider regional cooperation and integration initiatives (Chiang Mai Initiative Multilateralization, ASEAN capital market integration guidelines) in light of rising intra-regional connectedness.

*Sources: Haver Analytics and IMF staff calculations, as presented in the source content.*

### References ________________________________________________________________24

### _wp1657 - References ________________________________________________________________24

### Tables
- Table 1. Descriptive Statistics for Asset Return  _________________________________9
- Table 2. Descriptive Statistics for Daily Volatility _______________________________10

### Figures
- Figure 1. Time-varying Connectedness of Equity Returns and Volatilities_____________14
- Figure 2. Net Connectedness of Equity Returns: Japan, China, and Korea _____________15
- Figure 3. Dynamic Net Connectedness: Advanced Economies vs. Emerging Markets ____16
- Figure 4.    Net Directional Connectedness Index from China to Selected Countries______ 17
- Figure 5.    Net Connected Index from China_____________________________________17
- Figure 6. Connectedness Index – Equity Market Spillovers from China _______________19
- Figure 7. Total Connectedness Index: China to Hong Kong SAR  ___________________20
- Figure 8. Total Connectedness Index by Sector: China to Hong Kong SAR  ___________21

### Appendix
- Appendix Table 1. Equity Returns Connectedness  (Time Period: 06/30/2002–11/01/2015) _____________________________________________25
- Appendix Table 2. Equity Volatilities Connectedness (Time Period: 06/30/2002–11/01/2015) _____________________________________________26
- Appendix Figure 3. Dynamic Connectedness (Equity Returns and Volatility) _________27
- Appendix Figure 4. Dynamic Directional Connectedness of Equity Returns (Contributions to Others  _______________________________________________28
- Appendix Figure 5. Dynamic Directional Connectedness of Equity Volatility (Contributions to Others) __________________________________29
- Appendix Figure 6. Dynamic Directional Connectedness of Equity Returns (Contributions from Others)  ____________________________________________30
- Appendix Figure 7. Dynamic Directional Connectedness of Equity Volatility (Contributions from Others) ________________________________31
- Appendix Table 8. Dynamic Net Connectedness of Equity Return _________________32
- Appendix Table 9. Dynamic Net Connectedness of Equity Volatility _______________33

*https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1657.pdf*

### Appendix Table 10. Robustness Check: Dynamic Connectedness (Different Forecast

### Appendix Table 10. Robustness Check: Dynamic Connectedness (Different Forecast Horizons)

### I. Introduction and motivation
- Focus: quantify how shocks from Asian equity markets contribute to other countries’ equity returns and volatilities over time, and assess which markets are major sources and recipients of shocks.
- Context:
  - Financial integration within Asia has increased but remains relatively low: "only about 20 percent to 30 percent of cross-border portfolio investment and bank claims are intraregional."
  - Trade intraregional share: "53.3 percent of total Asia’s merchandise flows end up in Asia"; Euro area intra-regional trade share cited as "68.8%".
  - Example shock episode: China’s stock market fell sharply by "8.5 percent" on August 24, 2015; A-shares in the Shanghai Stock Exchange lost "one third" of their market value.
- Policy relevance: regional liquidity support arrangements (Chiang Mai Initiative Multilateralization) and ASEAN guidelines for capital market integration are ongoing; understanding connectedness informs policymakers and market monitoring during crises.

### II. Data and empirical setup
- Data:
  - Daily nominal local-currency stock market indexes from "January 1996 to October 2015" (Haver Analytics).
  - Sample: 13 Asian economies (Australia, China, Hong Kong SAR, India, Indonesia, Japan, Korea, Malaysia, New Zealand, the Philippines, Singapore, Taiwan Province of China, Thailand), four advanced countries (United States, United Kingdom, Germany, France) and three emerging markets from other regions (Brazil, Mexico, Turkey). Additional estimations included Europe and South Africa (with sample from "June 24, 2002").
- Returns and volatility construction:
  - Asset returns: change in the log-price, Friday-to-Friday (nominal return).
  - Volatility: Parkinson (1980) estimator using daily high and low prices; annualized daily volatility reported as percent standard deviation.
- Descriptive statistics provided for weekly returns (Table 1) and daily volatility (Table 2) in the source.

### III. Measures and methodology
- Connectedness framework: Diebold and Yilmaz connectedness measures (Diebold and Yilmaz 2009, 2012, 2014) based on forecast-error variance decompositions from VARs.
- Identification and advantages:
  - Use of the generalized VAR (Koop, Pesaran, and Potter 1996; Pesaran and Shin 1998) to obtain forecast-error variance decompositions invariant to variable ordering and allowing for correlated shocks.
- Key definitions:
  - Total connectedness: sum of cross-variance shares (fractions of H-step-ahead forecast-error variances of variable i due to shocks to variable j, j ≠ i).
  - Directional connectedness:
    - "From" connectedness: shocks received by vector i from all other vectors j.
    - "To" connectedness: shocks transmitted by vector i to all other vectors.
  - Net connectedness: difference between "To" and "From" connectedness for each market.
  - Dynamic connectedness: time-varying connectedness computed using rolling-window VAR estimates.
- Baseline estimation choices (reported):
  - VARs estimated with "four lags".
  - Connectedness calculated for a "10-week horizon".
  - Rolling-window baseline: "200-week window".
  - Robustness checks: different lags ("eight months"), forecasting horizons ("15, 20 weeks") and windows ("100 weeks").

### IV. Main empirical findings
- Aggregate connectedness trends:
  - "There has been a substantial increase in the aggregate equity return and volatility connectedness indexes since the global financial crisis."
  - Post-crisis: a mild decrease in recent years after a sharp rise in the aftermath of the global financial crisis.
- Returns versus volatilities:
  - Equity returns connectedness: smoother upward trending behavior.
  - Equity volatilities connectedness: displays spikes (burst behavior) in addition to the trend; both measures peaked during the global financial crisis.
- Advanced economies (AEs) vs. Emerging markets (EMs):
  - Since the global financial crisis, net connectedness indexes show a distinct pattern:
    - Advanced economies have increasingly become net "receiver" of shocks.
    - Emerging market economies, particularly China, ASEAN-5 (Indonesia, Malaysia, the Philippines, Singapore, and Thailand), and Korea, have increasingly become net sources of shocks.
  - Drivers:
    - For AEs: decrease in net connectedness driven by both a decline in "contribution to others" and an increase in "contributions from other countries".
    - For EMs: increase in net connectedness driven by an increase in "contribution to other economies".
- Role of China:
  - China’s net contribution to other markets rose, especially after the heightened volatility since summer 2015; China became a growing net source (transmitter) of shocks to other economies inside and outside the region, including effects on Hong Kong SAR, U.S., and European markets.
  - Contrasting trend: Japan’s contribution to other equity market movements declined over the same period.
- Channels and reasons for rising EM connectedness:
  - Direct financial linkages: growing weight of EMs in global portfolios; holdings of EM FDI and portfolio equity reached more than "15 percent of AE’s portfolio abroad, and 15 percent of AE’s GDP in 2012" (IMF 2014b citation in source).
  - Trade linkages: EMs’ economic developments exert greater influence via trade (e.g., China’s rebalancing affecting commodity prices).
  - Investor attention: international investors increasingly attentive to China as a destination of equity finance, increasing correlations of equity price changes between China and other Asian countries.

### V. Robustness and sensitivity
- Results robust to:
  - Different VAR lags (baseline four lags; mention of "eight months" alternative).
  - Different forecast horizons ("15, 20 weeks" tested in robustness).
  - Different rolling-window sizes ("100 weeks" alternative to baseline "200-week window").
  - Alternative orderings (generalized VAR chosen for ordering invariance; Cholesky decompositions also discussed for robustness checks).

### VI. Implications for monitoring and policy
- Monitoring:
  - Time-varying connectedness metrics (total, directional, net) are useful for tracking spillovers and for crisis monitoring, given their sensitivity to synchronized market events and spikes in volatility connectedness.
- Policy considerations:
  - Increasing role of EMs, especially China and several Asian EMs, as net shock givers suggests policymakers and market participants should factor emerging-market developments into global risk assessments.
  - Regional cooperation and integration initiatives (e.g., Chiang Mai Initiative Multilateralization, ASEAN capital market integration guidelines) are relevant given rising intra-regional connectedness, though financial integration still lags trade integration.
  - The changing composition of foreign banking claims (decline of euro area banks’ claims in emerging and developing Asia since 2008; expansion of Japanese and Chinese banks) is consistent with deeper regional financial integration and has implications for cross-border shock transmission.

*Source: Haver Analytics and IMF staff calculations, as presented in the source content.*

### 2015. The increased contribution of China’s equity returns to other markets is also far-

### _wp1657 - 2015. The increased contribution of China’s equity returns to other markets is also far-

### Rising connectedness of China’s equity market
- The net connectedness index (CI) from China to other equity markets has increased sharply since the second half of 2015.
- The average CI has increased since the global financial crisis in 2008 for most countries; in some cases there is a sharp rise since the second half of 2015.
- The steep increase in the CI in recent months is more pronounced than during the “taper tantrum” episode (May-July 2013), when the directional CI (from China) stayed relatively stable for nearly all economies considered.
- The sharp increase in CI since mid-2015 (relative to China’s equity market boom during mid-2014 to mid-2015) suggests that “contagion” effects—instead of direct financial linkages—could be at play.

### China’s influence across advanced and emerging markets
- The increased contribution of China’s equity returns extends to advanced economies (including the United States and euro area) and emerging economies.
- Figure labels indicate measured increases across groups: "Advanced Economies (All)", "Advanced Economies (Asia)", "Emerging Markets (All)", "Emerging Markets (Asia)" over the 1999–2014 timeline shown.
- Appendix statistics (Equity Returns Connectedness, Time Period: 06/30/2002-11/01/2015) show China row values including:
  - CHN row entries: 0.06, 0.06, 0.06, 0.07, 0.04, 0.07, 0.17, 0.01, 0.06, 0.06, 0.03, 0.03, 0.03, 0.02, 0.05, 0.03, 0.04, 0.02, 0.04, 0.05, From Others 83%, Net 0%
- Increase in Net Connectedness Index of Equity Returns from 2015 H2 to Jan 2016 listed across many economies (chart labels include USAGBREURAUSNZLSGPJPNHKGKORPHLIDNTHAMYSTWNBRAZAFTURINDMEX).

### China–Hong Kong SAR nexus and regional financial integration
- Hong Kong SAR is a regional financial hub; its ties to China have increased significantly over the last decade as China liberalized its financial markets and opened its financial account.
- Initiatives (e.g., Connect scheme) may have boosted financial integration and created new channels for transmission of volatility and financial stress.
- Methodology for Hong Kong SAR analysis:
  - VARs estimated with daily data and have four lags.
  - Spillovers calculated for a 10-week horizon.
  - Rolling-window estimations used to calculate time-varying CI and behavior during crisis and non-crisis periods.
  - Sample period covers June 2002–August 2015.
  - Sectoral indexes sample covers February 2005–August 2015.
  - Indexes used: Dow Jones Shanghai (China), Hang Seng Index (Hong Kong SAR), Hang Seng Composite Industry Index (Hong Kong SAR).
  - Baseline results based on a 200-week window; results robust to different lags (eight months), forecasting horizons (15, 20 weeks) and windows (100 weeks).
- Results:
  - Interlinkages between mainland China and Hong Kong SAR equity markets are strong and have been growing over time.
  - China’s equity markets explain a larger share of Hong Kong SAR’s equity returns (and volatility) than other economies.
  - Singapore’s average connectedness with the Mainland is similar to Malaysia and Korea (Figure 6).
  - Equity market spillovers from China to Hong Kong SAR have risen sharply for both returns and volatilities.
  - The return CI has been on the rise with a spike in late 2014; volatility CI reached a peak in 2011.
  - Net spillovers from China to Hong Kong SAR equity volatility (not shown) have been negative, suggesting Hong Kong SAR shocks have also hit China’s stock market.

### Sectoral heterogeneity in connectedness to Hong Kong SAR
- The increase in connectedness between China’s equity market and Hong Kong SAR’s stock market is broad-based but sectorally heterogeneous.
- Sectors with sharper rises in CI: Financials; Properties and Construction; Consumer goods.
- Information and Technology sector experienced a noticeable, albeit less drastic, increase.
- Financials and Properties and Construction have decoupled from other sectors, first from Information and Technology, and more recently from Consumer goods; possible reflection of stronger real economic linkages and financial exposures (e.g., Hong Kong SAR firms and banks’ exposure to China’s property markets).
- Figure timelines and scales shown for returns and volatilities (e.g., timeseries from 5/5/2006 to 5/5/2015 and 12/19/2008 to 12/19/2014) indicate CI magnitude up to at least 16 for sector indexes.

### Quantitative connectedness summary (selected totals)
- Appendix Table 1 (Equity Returns Connectedness, Time Period: 06/30/2002-11/01/2015):
  - "To others" totals by country (selected): USA To others 110%; GBR 118%; EUR 99%; CHN 84%; HKG 12%; KOR 108%; T WN 92%; SGP 64%; IND 84%; ZAF 49%; Total Connectedness (including own) example entries: USA 124%, GBR 130%, EUR 114%, CHN 100%, HKG 62%.
  - Individual country "From Others" and "Net" columns included (examples): USA From Others 78% Net 24%; GBR From Others 68% Net 30%; CHN From Others 58% Net 0%; HKG From Others 50% Net -38%.
- Appendix Table 2 (Equity Volatilities Connectedness, Time Period: 06/30/2002-11/01/2015):
  - "To others" totals by country (selected): USA 111%; GBR 133%; EUR 79%; CHN 49%; HKG 77%; SGP 130%; IND 102%; ZAF 67%; Total Connectedness (including own) example entries: USA 131%, GBR 150%, EUR 99%, CHN 68%, HKG 125%.
  - Individual country "From Others" and "Net" columns included (examples): USA From Others 68% Net 31%; GBR From Others 83% Net 50%; CHN From Others 48% Net -32%; HKG From Others 15% Net 25%; ZAF From Others 88% Net 74%.

### Robustness checks
- VAR specification variations tested:
  - Different windows and lags of the VAR system; main results derived from four-week lags, 10-week forecast horizons, and 100-week rolling window estimations; baseline uses 200-week window for dynamic CI.
  - Appendix Figure 10 presents dynamic connectedness index with different windows.
- Inclusion of exogenous global factors in VAR:
  - Exogenous variables tested: VIX, Citigroup Economic Surprise Index, Global commodity prices (S&P GSCI Spot Index), Bloomberg commodity spot index, monthly Global PMI (JP Morgan), monthly Global PMI Manufacturing only output (JP Morgan).
  - Exogenous factors entered separately and jointly.
  - Finding: results change little when exogenous global factors are included, allowing CI interpretation as shocks originating from each country netting out common global-factor shocks.

### Concluding remarks and policy implications
- Asian emerging markets, particularly China, have become important sources of financial shocks with increasingly large spillovers to other economies, including advanced market economies.
- China accounts for about one-third of global growth and has become a major player in global trade and increasingly, global finance.
- Since the global financial crisis, China has been increasingly a net “giver” of financial shocks to the rest of the Asia-Pacific region.
- Ongoing liberalization of China’s domestic financial markets and capital controls, and the internationalization of the RMB, suggest China’s importance to other Asian markets through financial channels will continue to rise.
- The rise in intraregional connectedness among emerging market economies confirms their role as a net source of shocks, expected to grow as trade and financial linkages strengthen.
- Policy recommendation:
  - Financial regulation and supervision should emphasize regional interlinkages and better understand gaps in regulation across markets and borders that may give rise to transmission channels for shocks.

*Sources: Haver Analytics and IMF staff calculations.*

### Appendix Figure 5.  Dynamic Directional Connectedness of Equity Volatility (Contributions to Others)

### Appendix Figure 5. Dynamic Directional Connectedness of Equity Volatility (Contributions to Others)

### Key findings (Contributions to Others by country — axis values and observed ranges)
- USA :  Contribution to Others — 80, 85, 90, 95
- GBR :  Contribution to Others — 84, 86, 88, 90, 92, 94
- EUR :  Contribution to Others — 60, 70, 80, 90, 100
- AUS :  Contribution to Others — 80, 85, 90, 95
- NZL :  Contribution to Others — 75, 80, 85, 90, 95, 100
- JPN :  Contribution to Others — 40, 50, 60, 70, 80, 90
- CHN :  Contribution to Others — 20, 40, 60, 80
- HKG :  Contribution to Others — 70, 75, 80, 85, 90, 95
- KOR :  Contribution to Others — 60, 70, 80, 90
- TWN :  Contribution to Others — 50, 60, 70, 80, 90, 100
- SGP :  Contribution to Others — 80, 85, 90, 95
- PHL :  Contribution to Others — 50, 60, 70, 80, 90
- IDN :  Contribution to Others — 40, 50, 60, 70, 80, 90
- THA :  Contribution to Others — 0, 20, 40, 60, 80, 100
- MYS :  Contribution to Others — 75, 80, 85, 90, 95, 100
- IND :  Contribution to Others — 40, 50, 60, 70, 80, 90
- BRA :  Contribution to Others — 40, 50, 60, 70, 80, 90
- MEX :  Contribution to Others — 60, 70, 80, 90, 100
- TUR :  Contribution to Others — 0, 20, 40, 60, 80, 100
- ZAF :  Contribution to Others — 75, 80, 85, 90, 95

*Sources: Haver Analytics and IMF staff calculations*

---

### Appendix Figure 6. Dynamic Directional Connectedness of Equity Returns (Contributions from Others)

### Key findings (Contributions from Others by country — axis values and observed ranges)
- USA :  Contribution from Others — 75, 80, 85, 90, 95
- GBR :  Contribution from Others — 80, 85, 90, 95
- EUR :  Contribution from Others — 80, 85, 90, 95
- AUS :  Contribution from Others — 75, 80, 85, 90, 95
- NZL :  Contribution from Others — 50, 60, 70, 80, 90
- JPN :  Contribution from Others — 70, 75, 80, 85, 90
- CHN :  Contribution from Others — 40, 50, 60, 70, 80
- HKG :  Contribution from Others — 75, 80, 85, 90, 95
- KOR :  Contribution from Others — 75, 80, 85, 90, 95
- TWN :  Contribution from Others — 70, 75, 80, 85, 90
- SGP :  Contribution from Others — 75, 80, 85, 90, 95
- PHL :  Contribution from Others — 50, 60, 70, 80, 90
- IDN :  Contribution from Others — 70, 75, 80, 85, 90
- THA :  Contribution from Others — 65, 70, 75, 80, 85, 90
- MYS :  Contribution from Others — 50, 60, 70, 80, 90
- IND :  Contribution from Others — 75, 80, 85, 90
- BRA :  Contribution from Others — 70, 75, 80, 85, 90
- MEX :  Contribution from Others — 75, 80, 85, 90, 95
- TUR :  Contribution from Others — 40, 50, 60, 70, 80, 90
- ZAF :  Contribution from Others — 75, 80, 85, 90, 95

*Sources: Haver Analytics and IMF staff calculations*

---

### Appendix Figure 7. Dynamic Directional Connectedness of Equity Volatility (Contributions from Others)

### Key findings (Contributions from Others by country — axis values and observed ranges)
- USA :  Contribution from Others — 60, 70, 80, 90
- GBR :  Contribution from Others — 65, 70, 75, 80, 85, 90
- EUR :  Contribution from Others — 65, 70, 75, 80, 85, 90
- AUS :  Contribution from Others — 60, 70, 80, 90, 100
- NZL :  Contribution from Others — 60, 70, 80, 90, 100
- JPN :  Contribution from Others — 40, 60, 80, 100
- CHN :  Contribution from Others — 50, 60, 70, 80, 90, 100
- HKG :  Contribution from Others — 50, 60, 70, 80, 90, 100
- KOR :  Contribution from Others — 70, 75, 80, 85, 90
- TWN :  Contribution from Others — 60, 70, 80, 90
- SGP :  Contribution from Others — 65, 70, 75, 80, 85, 90
- PHL :  Contribution from Others — 65, 70, 75, 80, 85, 90
- IDN :  Contribution from Others — 20, 40, 60, 80, 100
- THA :  Contribution from Others — 60, 70, 80, 90
- MYS :  Contribution from Others — 50, 60, 70, 80, 90
- IND :  Contribution from Others — 50, 60, 70, 80, 90
- BRA :  Contribution from Others — 60, 70, 80, 90
- MEX :  Contribution from Others — 60, 70, 80, 90, 100
- TUR :  Contribution from Others — 50, 60, 70, 80, 90, 100
- ZAF :  Contribution from Others — 70, 75, 80, 85, 90

*Sources: Haver Analytics and IMF staff calculations*

---

### Appendix Figure 8. Dynamic Net Connectedness of Equity Returns

### Key findings (Net Spillover by country — axis values and observed ranges)
- USA :  Net Spillover — 0, 5, 10, 15
- GBR :  Net Spillover — 0, 5, 10, 15
- EUR :  Net Spillover — -10, -5, 0, 5, 10
- AUS :  Net Spillover — 0, 5, 10, 15, 20
- NZL :  Net Spillover — 0, 5, 10, 15, 20, 25, 30
- JPN :  Net Spillover — -20, -15, -10, -5, 0, 5
- CHN :  Net Spillover — -30, -20, -10, 0, 10
- HKG :  Net Spillover — -5, 0, 5, 10
- KOR :  Net Spillover — -15, -10, -5, 0, 5
- TWN :  Net Spillover — -15, -10, -5, 0, 5
- SGP :  Net Spillover — 0, 2, 4, 6, 8, 10
- PHL :  Net Spillover — -10, -5, 0, 5
- IDN :  Net Spillover — -15, -10, -5, 0, 5
- THA :  Net Spillover — -10, -5, 0, 5
- MYS :  Net Spillover — 0, 5, 10, 15, 20
- IND :  Net Spillover — -15, -10, -5, 0, 5
- BRA :  Net Spillover — -20, -15, -10, -5, 0
- MEX :  Net Spillover — -4, -2, 0, 2, 4, 6
- TUR :  Net Spillover — -40, -30, -20, -10, 0
- ZAF :  Net Spillover — -5, 0, 5, 10, 15

*Sources: Haver Analytics and IMF staff calculations*

---

### Appendix Figure 9. Dynamic Net Connectedness of Equity Volatility

### Key findings (Net Spillover by country — axis values and observed ranges)
- USA :  Net Spillover — -10, 0, 10, 20, 30
- GBR :  Net Spillover — 0, 5, 10, 15, 20, 25
- EUR :  Net Spillover — -20, -10, 0, 10, 20
- AUS :  Net Spillover — -10, 0, 10, 20, 30
- NZL :  Net Spillover — -10, 0, 10, 20, 30, 40
- JPN :  Net Spillover — -30, -20, -10, 0, 10, 20
- CHN :  Net Spillover — -40, -30, -20, -10, 0, 10
- HKG :  Net Spillover — -10, -5, 0, 5, 10, 15
- KOR :  Net Spillover — -15, -10, -5, 0, 5
- TWN :  Net Spillover — -15, -10, -5, 0, 5, 10
- SGP :  Net Spillover — 0, 5, 10, 15, 20
- PHL :  Net Spillover — -20, -10, 0, 10, 20, 30
- IDN :  Net Spillover — -20, -15, -10, -5, 0, 5
- THA :  Net Spillover — -60, -40, -20, 0, 20
- MYS :  Net Spillover — 0, 10, 20, 30, 40
- IND :  Net Spillover — -30, -20, -10, 0, 10
- BRA :  Net Spillover — -30, -20, -10, 0, 10
- MEX :  Net Spillover — -10, -5, 0, 5, 10, 15
- TUR :  Net Spillover — -60, -40, -20, 0
- ZAF :  Net Spillover — -10, -5, 0, 5, 10

*Sources: Haver Analytics and IMF staff calculations*

---

### Appendix Figure 10. Robustness Check: Dynamic Connectedness (Different Forecast Horizons)

### Scenario description and observed total connectedness ranges
- Robustness check compares forecast horizons H=10 weeks (baseline) and H=20 weeks.
- Total Connectedness: H=10 weeks — 70, 72, 74, 76, 78, 80, 82, 84, 86, 88, 90
- Total Connectedness: H=20 weeks — 70, 72, 74, 76, 78, 80, 82, 84, 86, 88, 90

*Here, we tried different forecast horizons to measure the total connectedness index. H refers to the forecast horizon. The chart on the left is the baseline of 10 weeks. The chart on the right is using 20 weeks as a robustness check.*

*Sources: Haver Analytics and IMF staff calculations*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1657.pdf_