## _031114 — EXECUTIVE SUMMARY

## Source details

**Canonical URL:** [_031114 — EXECUTIVE SUMMARY](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2014/_031114.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2014/_031114.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2014/_031114.pdf.json)

---

### Context and purpose
- Date: March 11, 2014.
- Goal: Provide an operational framework for global financial conditions and global liquidity to inform Fund surveillance, building on prior IMF and BIS work.
- Motivation: Saint-Petersburg G20 Summit (September 2013) request to develop proposals to incorporate global liquidity indicators into Fund surveillance.

### Operational definition of global liquidity
- Global liquidity = the set of drivers of the supply of funds from international financial centers that affect the ease of global financing.
- Supply drivers include:
  - The nature and composition of investors.
  - Financial innovation.
  - General risk appetite.
  - Balance sheets of global financial and nonfinancial entities.
  - Policy settings in key economies, including prudential and financial regulation and monetary policy.

### Framework: drivers, transmission, outcomes
- Drivers (examples):
  - Prices: key policy rates, international money market and benchmark bond market rates.
  - Non-price factors: investors’ risk appetite, financial innovation, leverage.
  - Constraints: credit limits, margin requirements.
- Transmission channels:
  - Cross-border bank flows and non-core bank funding.
  - International portfolio flows.
  - International debt issuance and equity placements.
  - Operations of globally active financial intermediaries (G-SIFIs).
- Local outcomes:
  - Credit growth, cost of credit, credit conditions (including collateral and loan-to-value requirements), and asset prices.
  - Outcomes depend on domestic monetary stance, prudential policies, financial-sector structure, and external global liquidity.

### Relevance and policy considerations
- International financial conditions can provide support during stress but also introduce procyclical risks (asset price booms, buildup of vulnerabilities).
- Domestic monetary and fiscal policies can become less effective under strong global pressures.
- Macroprudential and prudential responses can influence vulnerability build-up.
- Monitoring global liquidity is important for bilateral and multilateral surveillance and to inform Fund recommendations.

---

### Key empirical findings and evidence of commonality
- Financial shocks propagate internationally and generate spillovers; co-movements in asset prices and quantities can exceed fundamentals.
- Synchronization:
  - Credit cycles across advanced economies are highly synchronized, with some increase over time.
  - Financial synchronization extends to many emerging markets; correlations of gross capital flows with the VIX are noted.
  - A latent-variable/common-factor approach can statistically account for much variability across countries.
- Quantitative concordance (Median Concordance Across Advanced OECD Countries):
  - Output: 80.27
  - House Prices: 59.25
  - Equity Prices: 71.26
  - Credit: 69.96
  - Sample periods referenced: Full Sample 1971-2011; Pre-Globalization 1971-84; Globalization 1985-2011 (table context provided).
- Illustrative factor correlations (with common factors):
  - Volatility (VIX): 0.12 and -0.10
  - US Credit Conditions Index: -0.55 and -0.60
  - Term spreads:
    - US Treasury 10/2y spread: -0.44 and -0.75
    - EA AAA Sov 10/5y spread: -0.59 and -0.93
    - UK Gilt 10/5y spread: -0.80 and -0.90
  - Policy rates:
    - US Fed Funds rate: 0.69 and 0.87
    - ECB MRO rate: 0.61 and 0.81
    - BOE OBR: 0.73 and 0.93

---

### Drivers and transmission: empirical insights

### What global factors drive cross-border bank flows?
- Confirmed explanatory variables:
  - VIX (volatility/uncertainty).
  - Global bank leverage.
  - U.S. yield curve slope.
  - M2 growth in advanced countries.
- Qualitative associations:
  - Flows decrease with greater volatility and term premia.
  - Flows increase with global bank leverage and money growth in advanced economies.
- Country-specific flows depend on global factors and country-specific characteristics.

### What global factors drive international portfolio flows?
- Asset manager presence expanded in bond markets: since 2008, fraction of bond markets under mutual fund management increased almost fivefold in EMs and threefold in AEs.
- Liquidity variables organized into 5 categories: (i) Price (term spread); (ii) Volatility; (iii) Quantity; (iv) Credit; (v) Monetary Policy.
- Associations:
  - Global liquidity–portfolio investment association stronger for bond flows than equity flows.
  - VIX and the TED spread are important drivers of portfolio flows.
  - Monetary policy conditions (especially Euro area and Japan) highly correlated with portfolio flows.
  - Credit conditions matter for bond flows.

### Common drivers (bank and portfolio flows)
- Cross-border banking and portfolio flows are affected by similar global drivers.
- Several global variables, including the VIX and the TED spread, co-move with both bank and portfolio flows.
- Some indicators (e.g., monetary aggregates) play a role in bank-credit transmission without being closely linked to bond and equity flow trends.
- Some variables affect bank flows in opposite directions to bond flows, indicating channel differences and potential substitution between bank and bond financing.

### Network centrality and transmission
- CPIS data show S4 economies are central in global portfolio claims networks (debt and equity); centrality reflects size and role as conduits for intermediation.
- Multilateral portfolio exposures have become somewhat more fragmented over the crisis, but the hub-and-spoke structure changed relatively slowly.

---

### Empirical results: coefficients, regressions, and explanatory power

### Global liquidity measure — core regression findings (Annex III)
- A one standard deviation in the global quantity measure implies:
  - a 2½ percent change in the quarterly change of cross-border claims on banks.
  - a 1½ percent change for non-banks.
- The summary global liquidity measure provides most explanatory power (about a 7 percentage points increase in R2, column 2 vs. column 1).
- Selected coefficients and model fit (from reported tables):
  - Panel A / Banks: coefficient on main macro control = 0.360*** (0.0580).
  - PCA_Quantity coefficients on global cross-border liquidity: 2.544*** (0.359); 2.535*** (0.347); 2.401*** (0.371); 1.917*** (0.387).
  - Alternative first-row coefficients: 0.360*** 0.168*** 0.359*** 0.167*** 0.165*** 0.182*** (standard errors in source).
  - Observations across reported columns: 2,715 2,254 2,254 2,254 2,254 2,254.
  - R-squared values reported: 0.038 0.106 0.039 0.106 0.106 0.118.
- Regression design notes:
  - Panel regressions with country fixed effects and clustered standard errors at the borrower country level.
  - Significance notation: ***,**, and * indicate statistical significance at the 1 percent, 5 percent, and 10 percent levels respectively. n.s. = non significant; (?) = inconclusive.

### Country policies and interactions (level and cyclicality effects)
- Level effects:
  - Capital controls associated with lower level of cross-border banking flows: capital controls coefficients reported: -0.0496** -0.0384* -0.0352** -0.0226.
  - Low institutional quality associated with lower cross-border flows: institution quality coefficients reported: -2.104*** -2.196*** -1.734*** -1.809***.
  - Restrictions on foreign bank entry reduce flows to both banks and non-banks: limits on foreign banks coefficients reported: -1.128*** 1.883** -0.819*** 0.400.
- Interaction with global liquidity (cyclicality):
  - Exchange rate flexibility * Global_CB_Liquidity: -0.630*** -0.196 (more flexible exchange rate reduces cyclicality of claims on banks).
  - Capital controls * Global_CB_Liquidity: -0.0307 -0.0194** (stricter capital controls reduce cyclicality of claims on non-banks).
  - Capital stringency * Global_CB_Liquidity: -0.207*** -0.0484 (greater capital stringency reduces cyclicality of claims on banks).
  - Supervisory power * Global_CB_Liquidity: -0.0936** -0.00438 (greater supervisory power reduces cyclicality of claims on banks).
  - Limits on foreign banks * Global_CB_Liquidity: -3.360*** -1.361*** (restrictions on foreign bank entry lower cyclical role of global liquidity).

### Portfolio flows: empirical coefficients and explanatory power (Empirical Specification / Table 2)
- Selected exact coefficients (preserved):
  - TED spread (3-month TED spread (LIBOR-Treasury)):
    - EM bond: 0.498***
    - AE bond: 0.512***
    - EM equity: 0.094***
    - AE equity: 0.157***
  - US Growth in Commercial Bank Loans:
    - EM bond: 0.355***
    - AE bond: 0.380***
    - EM equity: -0.003
    - AE equity: -0.196***
  - Japanese 10-year/3-months sovereign yield spread:
    - EM bond: 0.145***
    - AE bond: 0.213***
    - EM equity: 0.205***
    - AE equity: 0.260***
  - VIX (CBOE S&P 500 Volatility Index):
    - EM bond: 0.328***
    - AE bond: 0.230***
    - EM equity: 0.198***
    - AE equity: 0.243***
  - USA 10 year Off-the-run Liquidity Premium: EM bond 0.230***; AE bond 0.151***; EM equity 0.087***; AE equity 0.100***.
  - US Broker Dealer Leverage: EM bond -0.061***; AE bond -0.247***; EM equity 0.021*; AE equity 0.119***.
  - EA Deposit facility rate: EM bond 0.236***; AE bond 0.398***; EM equity 0.125***; AE equity 0.098***.
- R^2 and explanatory power (Table 3 / summary):
  - Bond flows:
    - EM: R^2 0.467
    - AE: R^2 0.465
  - Equity flows:
    - EM: R^2 0.233
    - AE: R^2 0.239
- Interpretation:
  - Association between global liquidity and portfolio investment is stronger for bond flows than for equity flows.
  - For bond flows, highest ranked price variable: TED spread; highest ranked credit variable: growth in US commercial bank loans.
  - For equity flows, price and volatility variables dominate: Japanese 10-year/3-months spread and the VIX.

---

### Analytics approach and selected annex findings

### Annex II — Common Component Analysis (methodology and results)
- Method:
  - Factor model separating price variables and quantity variables; extract first principal component for each cross-country sample as global price and quantity indices.
  - Balanced sample requirement: exclude series with less than eight years of data; remaining gaps imputed via iterated chained equations.
- Data:
  - Quantity: BIS Total Credit to the Private Sector, quarterly for 40 countries, expressed in real USD terms and (log) differenced.
  - Price: local interest rates, money-market rates, 10-year yields, commercial lending and mortgage rates, expressed in (real) levels.
  - Sample period: 1998Q4 to 2013Q2.
  - Price data sample dimensions: 59 observations over 112 variables.
  - Quantity data sample dimensions: 59 observations over 40 variables.
- Key statistical shares:
  - Price index represents around 45 percent of total variance of cross-country price data.
  - Quantity index presents almost 40 percent of the quantity-data variance.
- Bivariate supply–demand decomposition (Chen et al. methodology):
  - Identification: positive demand shock → price and quantity up; positive supply shock → price down and quantity up.
  - Main findings:
    - Expansion of supply helped keep prices down and potentially prolonged the 2002-08 boom.
    - Sudden contraction of supply associated with the crisis.
    - Recent efforts to boost supply are being offset by subdued demand.

### Annex VII — Banks’ noncore liabilities as indicators of vulnerability
- Two noncore indicator constructions:
  - Noncore 1: banks’ liability to the foreign sector + (M3 – M2).
  - Noncore 2: banks’ liabilities to the foreign sector + banks’ liabilities to the non-banking financial sector.
- Sample: 43 countries; timeframe: 1980 to first half of 2013.
- Estimation: panel probit with random effects; robustness: conditional logit with fixed effects.
- Main findings:
  - Noncore indicators are significant predictors of financial vulnerability, especially for banking crises (coefficients positive and mostly significant at 1 percent and 5 percent).
  - Predictive power is consistent across banking, stock market, and currency crises when noncore indicators are considered alone.
  - For banking crises, noncore indicators remain significant even after controlling for country-specific and global factors.
  - Inclusion of US noncore indicators (proxy for global liquidity) boosted model goodness-of-fit for banking crises from 60 percent to 90 percent, while removing statistical significance from local liquidity indicators.
  - For currency and stock market crises, noncore ratios’ predictive power weakens when country-specific macro variables and global risk appetite are controlled for.

### Annex VI — Off-shore debt issuance patterns (selected)
- Advanced economies remain main users of offshore debt, but EM presence has grown post-crisis.
- Largest increases in offshore debt issuance observed for China, with Brazil and Russia recording sizable increases.
- Sectoral composition examples (Dealogic / BIS stock as of 2013q3):
  - China: Dealogic stock 155 billions of USD — NFC 38%; Government 60%; Banks 0%; OFI 2%.
  - China: BIS stock 240 billions of USD — NFC 9%; Government 6%; Banks 26%; OFI 59%.
  - Brazil: Dealogic stock 248 billlions of USD — NFC 15%; Government 18%; Banks 30%; OFI 37%.
- Note: offshore financing for operations at home is sizable especially for China and Brazil; if intercompany lending is not included, currency mismatch risk may be understated.

### Annex VIII — Network representation of cross-border claims
- Network metrics built from CPIS bilateral claims; node centrality and link thickness represent value and relative importance.
- Key findings:
  - United States, United Kingdom, Euro area economies (France, Germany, Luxembourg), and Japan remain central in bond and equity portfolio networks.
  - Japan, Singapore and Hong Kong connect emerging Asia to the United States and Europe.
  - Network structure of cross-border debt and equity holdings changed little between 2006Q4 and 2012Q4, despite decline in magnitude of exposures.
  - China’s cross-border exposure with financial centers (Singapore, Hong Kong) increased significantly.

---

### Surveillance implications, dashboard, and policy use

### Dashboard approach and interpretation
- Convert variables into scores indicating deviation from historical mean, expressed in number of standard deviations.
- Large absolute scores warrant attention but do not automatically signal normative policy action.
- Stylized features identifiable with standardized variables:
  - Build-up of credit prior to the crisis.
  - Evaporation of liquidity during the crisis and surge in flows to safe havens (US, Germany).
  - Post-crisis flows to EMs; recent credit outcomes relatively subdued while bond and equity flows stronger amid accommodative monetary conditions including UMP.
  - Contrast: pre-crisis (low spreads, high policy rates) vs. current conditions (low policy rates, high spreads).

### Dashboard components (selected variables preserved verbatim)
- Drivers:
  - VIX
  - Broker Dealer Leverage = (Equity + Total Liabilities)/Equity
  - US Credit Manager's Index
  - US 10yr-2yr Spread (Percent)
  - Euro 10yr-3mth Spread (Percent)
  - Japan 10yr-3mth Spread (Percent)
  - US M2 (Annual Change (Percent))
  - Euro M2 (Annual Change (Percent))
  - Japan M2 (Annual Change (Percent))
  - US Policy Rate (Percent)
  - Euro Policy Rate (Percent)
  - Japan Policy Rate (Percent)
- Transmission:
  - Cross-Border Banking Flows (Annual Change (Percent), Exchange-Adjusted from 2013Q3 positions)
  - Portfolio Flows (EPFR Bond Funds Country Flows, Percent of Assets Under Management (AUM))
- Outcomes:
  - Total Credit (Real) = Total Credit to the Non-Financial Private Sector, all sources; USD, Deflated by Local CPI, Annual Change (Percent)
- Data sources cited: Haver Analytics; US Flow of Funds (US Federal Reserve); BIS Locational Banking Statistics, Table 6A; EPFR; BIS Total Credit Statistics.

### Integration with Fund surveillance and policy use
- Options:
  - Multilateral surveillance: incorporate liquidity conditions analysis in WEO projections, GFSR, and G-20 staff notes.
  - Bilateral surveillance: use spillover reports, Global-Risk Assessment Matrix, Vulnerability Exercises; focus on country-specific effects of global liquidity conditions.
- Considerations:
  - Liquidity spillovers can be beneficial but may render domestic countercyclical policies less effective or increase financial stability risks.
  - Country-specific circumstances and time variation require differentiated responses.

### Data limitations and next steps
- Data gaps: timeliness, high-frequency information on leverage and capital flows, detailed bilateral cross-border flows, and granular delineation of investor composition.
- SRFs database available for 142 countries provides consistent breakdown of non-core bank funding indicators; Fund working on more precise noncore liabilities measures using SRFs.
- Staff next steps:
  - Revisit and enrich conceptual foundations and empirical validity of the drivers/transmissions/outcomes framework, with focus on private liquidity creation in systemic economies and shadow banking links.
  - Further assess value of dashboard indicators.
  - Consider development and use of simple available indicators for systematic monitoring in bilateral and multilateral surveillance.

---

*Source document: _031114 — Global Liquidity: Issues for Surveillance (selected excerpts and annexes as provided in the supplied content).*

### EXECUTIVE SUMMARY

### _031114 - EXECUTIVE SUMMARY

### Context and Purpose
- March 11, 2014.
- Goal: Provide an operational framework for global financial conditions and global liquidity to inform Fund surveillance, building on prior IMF and BIS work.
- Motivation: The Saint-Petersburg G20 Summit (September 2013) called on the IMF “to carry out further research with a view to develop proposals on how to incorporate global liquidity indicators more broadly into the Fund's surveillance work.”

### Operational Definition of Global Liquidity
- Defined operationally as the factors that drive the supply of funding from international financial centers and thereby affect the ease of global financing.
- These factors include:
  - The nature and composition of investors.
  - Financial innovation.
  - General risk appetite.
  - Balance sheets of global financial and nonfinancial entities.
  - Policy settings in key economies, including prudential and financial regulation and monetary policy.

### Framework: Drivers, Transmission, Outcomes
- Drivers (examples):
  - Prices: key policy rates, international money market and benchmark bond market rates.
  - Non-price factors: investors’ risk appetite, financial innovation, leverage.
  - Constraints: credit limits, margin requirements.
- Transmission channels:
  - Cross-border bank flows and non-core bank funding.
  - International portfolio flows.
  - International debt issuance and equity placements.
  - Operations of globally active financial intermediaries (G-SIFIs).
- Local outcomes:
  - Credit growth, cost of credit, credit conditions (including collateral and loan-to-value requirements), and asset prices.
  - Outcomes depend on domestic monetary stance, prudential policies, financial-sector structure, and external global liquidity.

### Key Empirical Findings and Evidence of Commonality
- Financial shocks propagate internationally and generate spillovers; co-movements in asset prices and quantities can exceed fundamentals.
- Synchronization:
  - Credit cycles across advanced economies are highly synchronized, with some increase over time.
  - Financial synchronization extends to many emerging markets, with correlations of gross capital flows with the VIX noted.
  - A latent-variable/common-factor approach can statistically account for much variability across countries.
- Quantitative concordance (Median Concordance Across Advanced OECD Countries):
  - Output: 80.27
  - House Prices: 59.25
  - Equity Prices: 71.26
  - Credit: 69.96
  - Sample periods shown: Full Sample 1971-2011; Pre-Globalization 1971-84; Globalization 1985-2011 (table context provided).
- Factor correlations (illustrative correlations with common factors):
  - Volatility (VIX): 0.12 and -0.10
  - US Credit Conditions Index: -0.55 and -0.60
  - Term spreads:
    - US Treasury 10/2y spread: -0.44 and -0.75
    - EA AAA Sov 10/5y spread: -0.59 and -0.93
    - UK Gilt 10/5y spread: -0.80 and -0.90
  - Policy rates:
    - US Fed Funds rate: 0.69 and 0.87
    - ECB MRO rate: 0.61 and 0.81
    - BOE OBR: 0.73 and 0.93

### Analytics: Drivers and Channels (Empirical Insights)
- Empirical analysis links price and quantity factors—measures of risk, bank leverage, and interest rates in financial centers—to cross-border bank flows and international portfolio flows.
- Country-specific policies, including exchange rate and prudential frameworks, shape how global conditions are transmitted.
- Significant unknowns remain: evolving global funding structures, changing institutions, and financial innovation affect liquidity creation mechanics and transmission channels.

### Surveillance Implications and Indicators
- The paper suggests indicators across various types of economies for tracking global liquidity, organized into a dashboard tracing trends over time.
- The dashboard:
  - Tracks evolution of global financial conditions.
  - Includes several indicators already monitored in flagship multilateral surveillance products.
- Limitations and needs:
  - State of the art is limited; continuous review of indicators is warranted given changing institutional environments, regulatory policies, and financial innovation.
  - A better understanding of policy drivers in advanced economies and international propagation mechanisms is needed.

### Relevance and Policy Considerations
- International financial conditions can provide support during stress but also introduce procyclical risks (asset price booms, buildup of vulnerabilities).
- Policy effectiveness:
  - Domestic monetary and fiscal policies can become less effective under strong global pressures.
  - Macroprudential and prudential responses can influence vulnerability build-up.
- Monitoring global liquidity is important for both bilateral and multilateral surveillance and can help integrate policy advice with implications for Fund recommendations.

### Analytics Approach and Annexes (structure overview)
- Factor analysis distinguishing quantity and price dimensions of credit across 40 countries to identify common trends and to help disentangle supply vs. demand contributions to credit outcomes.
- Annexes and boxes cover: definitions of liquidity, common component analysis, global liquidity and cross-border bank lending, international asset management, portfolio capital flows, off-shore debt issuance patterns, banks’ noncore liabilities, and network representation of cross-border claims.

*Prepared by an interdepartmental IMF team; approved by Stijn Claessens, Gaston Gelos, Vikram Haksar.*

### 8.      Global liquidity, booms and busts. The run up to the global financial crisis provides the

### 8.      Global liquidity, booms and busts. The run up to the global financial crisis provides the

### A. Context and main narrative
- The 2003-08 run up to the global financial crisis exemplifies how financial conditions in the S4 appear to have driven 2003-08 global conditions: loose monetary stance, steep yield curve, and financial innovation facilitated increased leverage and expanded intermediary balance sheets.
- As balance sheets grew and marked-to-market equity capital rose, lenders’ incentives led to looser credit standards (e.g. subprime mortgage), with reinforcing feedback among lower risk measures, expanding balance sheets, and higher asset valuation.
- Post-crisis cross-border deleveraging of U.S. and European banks and capital markets reverberated globally.
- Large-scale monetary policy response in the S4 helped buffer the unwinding of private liquidity in advanced markets and had global implications; unconventional monetary policies (UMP) and their global impact have been subject to attention (see IMF (2013a, 2013c)).

### B. An operational definition of global liquidity
- Global liquidity is not yet well understood; a comprehensive international treatment remains work in progress.
- Operational definition: global liquidity = the set of drivers of the supply of funds from international financial centers (conditions in the S4 financial centers largely determine ease of global financing and influence international financial flows to other AEs and EMs).
- Framing: shifts in the supply function for cross-border funding determine global liquidity; equilibrium price (expected return differentials, adjusted for country specific risks) balances supply and demand, but willingness to provide funding is also determined by non-price supply factors.

### C. Candidate supply factors for cross-border funding (empirical correlates listed in the text)
- Risk attitudes of major investors, including risk off/risk on episodes, often proxied by the VIX and other uncertainty indicators.
- Balance sheet conditions of global banks (leverage: accounting or market-based indicators) and the size/conditions of the shadow banking system.
- The general (as opposed to relative) level of interest rates, including yield curve slope, through risk taking.
- Monetary policy in S4 economies and some money aggregates such as M2 (link to supply of funds unclear but probably via banks’ liabilities).
- Macroprudential policies and financial (de-)regulations in origin and destination countries; financial innovation.
- Composition of investor pools: banks, real money investors, hedge funds, domestic versus foreign residents.

### D. Empirical agenda and preliminary findings
- Proposed empirical agenda: assess role of each candidate factor, with focus on S4 policy settings, interest rate differentials, banks’ balance sheets, and effects on risk off/risk on conditions.
- BIS locational and balance sheet data show a few advanced economies dominate cross-border banking; cross-border lending patterns show heterogeneity by destination but strong commonalities over time.
- Empirical exercises use BIS locational data and EPFR Global mutual fund flows (normalized by AUM) to study bank and portfolio channels.

### E. What global factors drive cross-border bank flows?
- Empirical results confirm explanatory power of:
  - VIX (volatility/uncertainty)
  - Global bank leverage
  - U.S. yield curve slope
  - M2 growth in advanced countries
- Qualitative findings:
  - Flows decrease with greater volatility and term premia.
  - Flows increase with global bank leverage and money growth in advanced economies.
- Country-specific flows depend on global factors and country-specific characteristics (Annex III referenced).

### F. What global factors drive international portfolio flows?
- Asset managers and mutual funds have expanded global presence, especially in bond markets: since 2008, fraction of bond markets under management by mutual funds increased almost fivefold in EMs and threefold in AEs.
- Analysis focuses on S4 financial condition indexes (e.g., Chicago Fed’s NFCI for the United States) and constructs analogous measures for Euro Area, Japan, United Kingdom.
- Liquidity variables organized into 5 categories: (i) Price (term spread); (ii) Volatility; (iii) Quantity; (iv) Credit; (v) Monetary Policy.
- Empirical ranking within categories identifies best-performing variables; results robust to alternate specifications and subsamples (2003–Aug-2008; Sep-2008 onwards).
- Associations:
  - Global liquidity–portfolio investment association stronger for bond flows than equity flows.
  - VIX and the TED spread are important drivers of portfolio flows.
  - Monetary policy conditions (especially Euro area and Japan) highly correlated with portfolio flows.
  - Credit conditions matter for bond flows.

### G. Network centrality and transmission
- CPIS data show S4 economies are central in global portfolio claims networks (debt and equity); centrality reflects size and role as conduits for intermediation.
- Multilateral portfolio exposures have become somewhat more fragmented over the crisis, but hub-and-spoke structure has changed relatively slowly.

### H. Common drivers: bank and portfolio flows
- Cross-border banking and portfolio flows are affected to a high degree by similar global drivers.
- Several global variables, including the VIX and the TED spread, co-move with both bank and portfolio flows.
- Some indicators (e.g., monetary aggregates) play a role in bank-credit transmission without being closely linked to bond and equity flow trends.
- Some variables affect bank flows in opposite directions to bond flows, suggesting different channels and substitution effects between bank and bond financing.

### I. Summary of empirical results (as presented in Table 2)
- Risk appetite
  - VIX: negative association with Bank flows to banks, Bond flows to non-banks, Bond flows to AE, Bond flows to EM, Equity flows to AE, Equity flows to EM (indicated as VIX (-) *** across columns).
  - TED spread: Bank flows to banks (-) *; Bond flows to non-banks n.s.; Bond flows to AE (-) ***; Bond flows to EM (-) ***; Equity flows to AE (-) ***; Equity flows to EM (-) ***.
- Balance sheet conditions
  - Broker dealer leverage: Bank flows to banks (+) ***; Bond flows to non-banks (+) ***; Bond flows to AE (-) ***; Bond flows to EM (-) ***; Equity flows to AE (+) ***; Equity flows to EM (+) *.
- Credit conditions in the U.S.
  - Findings inconclusive for bank flows (?), but positive for Bond flows to AE (+) ***, Bond flows to EM (+) ***, Equity flows to AE (+) ***, Equity flows to EM (+) ***.
- Interest rates
  - Yield curve slope: Bank flows to banks (-) ***; Bond flows to non-banks (-) ***; Bond flows to AE (+) ***; Bond flows to EM (+) ***; Equity flows to AE (?) ; Equity flows to EM (?).
- Monetary policy
  - Policy rates: Bank flows to banks (+) ***; Bond flows to non-banks (+) ***; Bond flows to AE (-) ***; Bond flows to EM (-) ***; Equity flows to AE (?) ; Equity flows to EM (?).
  - M2: Bank flows to banks (+) ***; Bond flows to non-banks (+) **; Bond flows to AE n.s.; Bond flows to EM n.s.; Equity flows to AE n.s.; Equity flows to EM n.s.
- Notes on Table 2 notation: (+) and (-) indicate a significant coefficient with positive or negative sign respectively. ***,**, and * indicate statistical significance at the 1 percent, 5 percent, and 10 percent levels respectively. n.s. stands for non significant, and (?) indicates inconclusive findings (e.g. different signs across specifications).

### J. Policy-relevant interactions between global drivers and recipient-country policies
- Country policies can mitigate cyclicality of cross-border flows:
  - More flexible exchange rate regime reduces cyclicality of cross-border banking flows to bank borrowers.
  - Stricter capital controls reduce cyclicality of cross-border banking flows to non-bank borrowers.
  - Greater capital stringency and/or more supervisory power reduce cyclicality of cross-border claims on banks (but not claims on non-banks).
  - Restrictions on foreign bank entry lower the cyclical role of global liquidity factors in general.
- Other findings:
  - Presence of capital controls and low quality institutions are associated with lower cross-border flows.
  - Restrictions on foreign-bank entry or ownership reduce flows to both banks and non-banks.
- For portfolio flows:
  - Institution quality matters for both bond and equity portfolio flows.
  - Exchange rate flexibility reduces impact of global drivers on equity portfolio flows.
  - Bond flows are reduced by capital controls; equity flows are reduced by limits on foreign bank presence.

### K. Changing patterns (selected illustrations referenced)
- Shifts in channels of external debt financing evident in cumulative BoP portfolio debt liability flows (banks vs non-bank corporate vs EM).
- Central bank balance sheet expansions shown via ratios: Eurosystem total assets and Fed total assets relative to GDP compared with bank liquidity measures (core vs non-core).

*Italic: Source document: _031114 - 8.      Global liquidity, booms and busts. The run up to the global financial crisis provides the*

### 25.      Continuous shifts in international banks’ funding structure are an essential aspect of

### Continuous shifts in international banks’ funding structure are an essential aspect of

### Non-core funding and private liquidity
- Global banks’ increased reliance on “non-core” (wholesale) funding was a key aspect of the surge in balance sheets and cross-border lending in the pre-crisis (2003-07) phase.
- The increase in non-core funding of banks in smaller open economies indicated easier access to a growing pool of cross-border liquidity (with associated vulnerabilities, see Annex VII).
- The collapse in non-core funding was one aspect of the sharp contraction in private liquidity that motivated the extension of central bank liquidity provision.
- Within the Euro area, increased reliance on Eurosystem long term refinancing operations substituted for cross-border interbank lending to crisis-affected countries.

### Public liquidity provision has buffered private deleveraging
- Official liquidity is usually a small part of the financial system in advanced economies, but unconventional monetary policies (UMP) were warranted after the Lehman crisis and the Euro area crisis to buffer net impacts on financial conditions.
- Central bank actions offset aspects of private deleveraging through large-scale policy measures (including long-term refinancing operations in the Euro area).

### Patterns of global finance since the crisis: debt issuance and offshore activity
- Patterns have shifted toward more reliance on debt issuance, especially by EM corporations.
- The upward trend in international debt issuance that halted in 2008 had primarily benefitted advanced economies; since 2010 EM international debt issuance has increased.
- Regulatory changes (e.g., transition to Basel 3) may have shifted intermediation from banks to capital markets.
- Notable features:
  - Access gained by “frontier markets” to international funding through debt issuance.
  - EM international debt issuance increasingly taking place off-shore (see Box 3).
- Box 3 highlights:
  - AEs’ reliance on off-shore debt surged pre-crisis and has since declined as EMs took a larger share.
  - Issuance out of off-shore financial centers stabilized and even increased since 2010.
  - The largest volume increase in offshore debt issuance has been for China, with Brazil and Russia recording sizable increases.
  - Offshore financing for operations at home is sizable especially for China and Brazil; balance of payments statistics indicate funds raised by offshore subsidiaries returned as reverse investment (FDI).
  - If intercompany lending is not included, currency mismatch risk may not be fully captured in external debt vulnerability assessment.
  - Increased offshore usage by EMs suggests a need to trace how this liquidity enters EM domestic financial systems and whether it creates new vulnerabilities.
  - Chung et al. (2014) suggest offshore foreign currency borrowing by firms may turn up in corporate local currency deposits, motivating monitoring of this monetary data.

### Liquidity indicators and Fund surveillance: drivers, transmission, outcomes
- Recommended monitoring focuses on:
  - Key drivers: policy rates, yield curve configurations, size of balance sheets in major funding economies.
  - Transmission: intermediaries’ balance sheets, cross-border flows.
  - Outcomes: country-specific credit outcomes.
- There is a need to monitor multiple indicators because drivers provide different incentives across agents and evolve with market structures.
- Data constraints and access issues limit monitoring (see Box 4). For example, several countries reporting to the BIS International Banking Statistics limit the Fund’s access to the full restricted dataset.
- Box 4 (data efforts) notes:
  - SRFs database available for 142 countries provides consistent and detailed breakdown of non-core bank funding indicators (as used in Annex VII).
  - SRFs includes local currency versus foreign currency breakdowns, and detailed sector and nonresident sector breakdowns useful for assessing currency and liquidity crisis risks.
  - The Fund is working on developing more precise measures of noncore liabilities based on the SRFs database.
  - Remaining gaps include timeliness, high frequency information on leverage and capital flows, detailed bilateral cross-border flows, and a more granular delineation of the composition of international investors.
  - A closer examination of the balance sheet structure of large EM financial intermediaries would be valuable.

### Dashboard approach and interpretation
- Approach: convert variables into scores indicating deviation from historical mean, expressed in number of standard deviations.
- Large absolute scores warrant attention but do not automatically signal normative policy action (policy variables may need large deviations to buffer private liquidity dynamics).
- Stylized features identified by standardized variables:
  - Build-up of credit prior to the crisis.
  - Evaporation of liquidity during the crisis and surge in flows to safe havens (US, Germany).
  - Post-crisis flows to EMs; recent credit outcomes relatively subdued while bond and equity flows have been stronger amid accommodative monetary conditions including unconventional policies.
  - Contrast between pre-crisis global factors (low spreads, high policy rates) and current conditions (low policy rates, high spreads) highlights need for nuanced analysis.

- Dashboard components (selected variables and definitions preserved from source):
  - Drivers: VIX; Broker Dealer Leverage = (Equity + Total Liabilities)/Equity; US Credit Manager's Index; US 10yr-2yr Spread (Percent); Euro 10yr-3mth Spread (Percent); Japan 10yr-3mth Spread (Percent); US M2 (Annual Change (Percent)); Euro M2 (Annual Change (Percent)); Japan M2 (Annual Change (Percent)); US Policy Rate (Percent); Euro Policy Rate (Percent); Japan Policy Rate (Percent).
  - Transmission: Cross-Border Banking Flows (Annual Change (Percent), Exchange-Adjusted from 2013Q3 positions); Portfolio Flows (EPFR Bond Funds Country Flows, Percent of Assets Under Management (AUM)).
  - Outcomes: Total Credit (Real) = Total Credit to the Non-Financial Private Sector, all sources; USD, Deflated by Local CPI, Annual Change (Percent).
  - Sources cited for variables: Haver Analytics; US Flow of Funds series published by the US Federal Reserve; BIS Locational Banking Statistics, Table 6A; EPFR; BIS Total Credit Statistics.

### Integration with Fund surveillance and policy use
- Options for integrating indicators:
  - Multilateral surveillance: incorporate liquidity conditions analysis in WEO projections, the GFSR for high frequency surveillance, and G-20 staff notes.
  - Bilateral surveillance: use spillover reports and the Global-Risk Assessment Matrix; integrate aspects into Vulnerability Exercises; focus on country-specific effects of global liquidity conditions.
- Considerations:
  - Liquidity spillovers may be beneficial for recipients; however, they may render domestic countercyclical policies less effective or increase financial stability risks, with potential welfare costs.
  - Country-specific circumstances and time-varying factors require differentiated policy responses.

### Next steps and research agenda
- Staff next steps include:
  - Revisiting and enriching the conceptual foundations and empirical validity of the drivers/transmissions/outcomes framework, with particular interest in private liquidity creation in systemic economies and links to shadow banking (staff will probe further, including in the Fall 2014 GFSR).
  - Further assessing the value of monitoring indicators such as those in the "dashboard".
  - Considering development and use of simple, available indicators to support more systematic and regular monitoring of global financial conditions in staff’s bilateral and multilateral surveillance.

*Source: IMF chapter section as provided in the supplied content.*

### Chapter 12, pp.601-650.

### Chapter 12, pp.601-650

### Scope and limitations
- The state of knowledge on the drivers of global liquidity, defined as the “ease of international financing” is limited.
- Factors driving the ease of financing globally and how they affect countries are not well understood.
- Fuller understanding will emerge at the intersections and integration of multiple research areas, some well established and others recently explored (Landau (2013), Shin (2012), CGFS (2011), Obstfeld (2009); Holmstrom and Tirole (2011)).

### Financial accelerator (demand-side propagation)
- Core mechanism: shocks (asset prices or real economic prospects) influence borrowers’ balance sheets; financial imperfections cause changes in borrowers’ net worth to affect the volume and cost of external finance.
- These propagation mechanisms can have general equilibrium effects and lead to procyclicality (Bernanke and Gertler (1989); Carlstrom and Fuerst (1997); Kiyotaki and Moore (1997); Bernanke, Gertler and Gilchrist (1999)).
- In open-economy contexts:
  - Financial imperfections are likely stronger because contracts are harder to enforce and information asymmetries are greater across borders.
  - Pledgeability of output and verifiability of borrowers’ quality are more limited, so net worth has a larger influence on access to finance.
  - As exchange rates and asset prices increase and countries’ net worth rises, external financing volume tends to increase and cost tends to decline; the reverse occurs when net worth falls.
  - These channels are quantitatively important in explaining procyclical external financing and the synchronized nature of global crises (Gertler, Gilchrist and Natalucci (2007); Perri and Quadrini (2010)).

### Supply-side of finance: bank balance sheets and the bank capital channel
- Emphasizes the role of banks and other financial intermediaries in providing financing and liquidity, determining asset prices, and affecting the real economy.
- Changes in an intermediary’s net worth affect its access to and cost of funding and its ability to make new loans and undertake intermediation.
- The bank capital channel: weakened loan portfolio quality during recessions reduces bank capital; unable to raise capital quickly, banks tighten lending standards and reduce risky credit supply, producing a “credit crunch” and deeper downturns (Van den Heuvel (2008)).
- When many banks are hit by the same shock, economy-wide effects can arise.

### Leverage cycles, search-for-yield, and endogenous supply-side dynamics
- Endogenous leverage cycles: when measured capital is high and perceived risk is low, intermediaries’ balance sheets expand and leverage increases (Adrian and Shin (2010, 2011)).
- Low interest rates and high asset prices make these dynamics more likely; monetary policy stance can influence private financing provision.
- A “search for yield” can occur among banks and institutional investors when interest rates are relatively low (Dell’Ariccia and Marquez (2013); Borio and Zhou (2008)).
- Rapid shifts in equilibria and non-linearities can lead to abrupt reductions in liquidity and external financing; parallels to self-fulfilling crises exist (Obstfeld (1986); Geanakoplos (2010)).

### International implications and cross-border channels
- Supply-side channels (bank capital, leverage) are highly relevant internationally even if some domestic channels (e.g., small firms’ access to cross-border loans) are less so.
- Empirical evidence:
  - International banks’ pullback from cross-border lending after the recent financial crisis illustrates the capital channel, with the degree of pullback depending on balance sheet impact (Cerutti and Claessens, 2014).
  - Bruno and Shin (2013a, 2013b) model capital and leverage channels generating procyclical international lending and demonstrate empirical relevance for cross-border banking flows.
  - Hahm, Shin, and Shin (2013) show reliance on non-core funds predicts banking system vulnerabilities by proxying ease of funding conditions.
  - Chung et al. (2014) show non-core liabilities of international banks signal broad credit conditions and predict global trade and growth; Shin (2013) is cited for related analysis.
  - Rey (2013) attributes high co-movements in international financial markets to similarities in funding conditions and balance-sheet positions of internationally active banks and intermediaries in key financial centers.

### Synthesis: interaction of demand- and supply-side channels
- Global liquidity dynamics reflect interactions between:
  - Demand-side financial accelerator effects that amplify shocks via borrower net worth and access to external finance.
  - Supply-side bank capital, leverage, and intermediation dynamics that determine funding availability, cost, and credit supply.
- Monetary policy, funding conditions, and balance-sheet positions of intermediaries together shape cross-border financing cycles and the international transmission of shocks.

*Source: Chapter 12, pp.601-650.*

### Annex II. Common Component Analysis

### Annex II. Common Component Analysis

### Methodology
- Factor model applied to a broad set of credit indicators, based on Stock and Watson (2002), Bai and Ng (2002) and Eickmeier and others (2013).
- Objective: explore global commonality in dynamics of indicators, defined as the share of variance explained by a common factor estimated via factor analysis.
- Steps:
  - Combine a large set of credit indicators across countries into a single cross-country dataset; each indicator is centered and standardized.
  - Separate data into i) price variables, and ii) quantity variables.
  - Extract the first principal component from each separate cross-country sample.
  - Use the latent first component as the common “global” portion of cross-country liquidity, serving as single indicators of either the price or quantity of global liquidity.
- Data handling and sample requirements:
  - Principal Components Analysis requires a balanced sample; exclude all series with less than eight years of data.
  - Any remaining data gaps are imputed using iterated chained equations across the full dataset (White and others, 2011).
- Note: Analysis aligns with factor-analysis approach (explicitly dividing price- and quantity-related variables and taking the first factor as the global common trend).

### Data
- Quantity measure:
  - BIS data for Total Credit to the Private Sector, quarterly for 40 countries (advanced and emerging).
  - Covers total credit provided by domestic banks, all other sectors, and non-residents.
  - Quantity variables expressed in real USD terms, and are (log) differenced.
- Price measures:
  - Range of local domestic interest rates including policy rates; money-market rates; 10-year government bond yields; commercial lending rates; mortgage rates.
  - Interest-rate variables are expressed in (real) levels.
- Sample period: 1998Q4 to 2013Q2.
- Robustness checks and extensions:
  - Alternative quantity indicator: cross-border bank credit flowing into a country from BIS locational data (exchange-adjusted).
  - Broader range of local quantity-related variables: M0 growth; M2 growth; cross-border credit growth; domestic-currency loans from local affiliates (nominal framing).
  - Include additional global factors: VIX and US term spreads in price analysis; US repo volume growth and commercial-paper growth in quantity analysis.
- Sample dimensions (as reported):
  - The price data includes 59 observations over 112 variables.
  - The quantity data includes 59 observations over 40 variables.
- Prepared by: Andrew Tiffin (SPR).

### Results
- Interpretation: Each extracted first component is a standardized index; captures main features of global liquidity over the past decade.
- Key statistical shares:
  - The single price index represents around 45 percent of the total variance of the cross-country price data.
  - The quantity index presents almost 40 percent of the quantity-data variance.
- Subsample analysis:
  - Price and quantity indexes calculated over two subsamples (S4 countries and non-S4 countries) are broadly similar to total-sample counterparts, suggesting the two groups were generally coupled over the sample period.
- Qualitative features of indexes:
  - Quantity series:
    - Captures contraction following the 2001 dot-com bust, long easing over the mid-2000s, sharp drop following the 2008 crisis.
    - Current index suggests credit remains relatively subdued.
  - Price series:
    - Covers easing of policy rates after dot-com bust, subsequent increase after 2004.
    - Crisis-period developments dominated by impact of inflation on real financing conditions: surge in oil and commodity prices during 2002-2008 boom (particularly 2007) raised CPI and effectively eased real financing conditions.
    - Following the crisis, drop in commodity prices and inflation contributed to effective tightening of real rates.
    - Currently, global funding costs remain relatively low; but the recent downward trend in inflation has produced a slight tightening in real terms.
- Robustness checks (Figure 2 summary):
  - Different quantity measures capture same broad patterns.
  - The final indicator based on nominal data does not display the same inflation-based swings in prices immediately before and after the global financial crisis.

### Extension: Supply and Demand
- Identification strategy:
  - Use indices for price and quantity and apply methodology from Chen and others (2012) to decompose movements into supply and demand shocks.
  - Model detrended indexes in a bivariate vector autoregression (VAR); regress detrended indexes on their own lagged values and a constant.
  - From reduced-form residuals, construct structural innovations that are serially and contemporaneously uncorrelated.
  - Impose sign identification restrictions:
    - Positive demand shock → increase in both price and quantity of credit.
    - Positive supply shock → fall in price and increase in quantity.
  - Structural identification details:
    - Translate reduced-form VAR into structural model via inverse of Cholesky decomposition.
    - Any matrix (PQ)^{-1}, where P is Cholesky and Q is orthonormal, yields observationally equivalent models; choose Q to match identification restrictions.
    - Use median targeting approach to select a specific model: choose model with impulse responses as close as possible to the median response among 10,000 randomly selected models (Fry and Pagan, 2010).
  - Use historical (Wold) decomposition to outline cumulative contributions of supply and demand to both price and quantity indices.
- Main findings from decomposition:
  - Expansion of supply helped keep prices down and potentially prolonged the 2002-08 boom.
  - Sudden contraction of supply associated with the crisis.
  - Recent efforts to boost supply (perhaps through expansionary policy) are currently being offset by subdued demand.
- Caveats:
  - Results subject to usual caveats given variation of individual-country data around estimated components and abstract nature of components.

*Source: Annex II. Common Component Analysis (prepared by Andrew Tiffin, SPR) — IMF staff analysis as presented in the supplied content.*

### Annex III) in Table 2, we confirm that this global liquidity measure is a statistically significant

### Annex III) in Table 2, we confirm that this global liquidity measure is a statistically significant

### Global liquidity measure — core findings
- A one standard deviation in the global quantity measure implies a 2½ percent change in the quarterly change of cross-border claims on banks (and 1½ percent for non-banks).
- The summary global liquidity measure provides most explanatory power (about a 7 percentage points increase in R2, column 2 vs. column 1), more than any of the other variables.
- Only the VIX provides some additional explanatory power (column 6; the regression results with the other variables are not reported).

### Key regression coefficients and model fit (selected from Table 2 and Table 1)
- Panel A / Banks: coefficient on main macro control (first listed) = 0.360*** (0.0580).
- Panel A / Banks: other baseline coefficient = -0.116*** (0.0288).
- Table 2: summary (PCA_Quantity) coefficients on global cross-border liquidity:
  - 2.544*** (0.359)
  - 2.535*** (0.347)
  - 2.401*** (0.371)
  - 1.917*** (0.387)
- Table 2: alternative reported coefficients (first row across columns) = 0.360*** 0.168*** 0.359*** 0.167*** 0.165*** 0.182*** (standard errors in parentheses in source).
- Table 2: Observations = 2,715 2,254 2,254 2,254 2,254 2,254 (as reported across columns).
- Table 2: R-squared = 0.038 0.106 0.039 0.106 0.106 0.118 (as reported across columns).
- Notes: regressions are panel regressions with country fixed effects and clustered standard errors at the borrower country level; significance: *** p<0.01, ** p<0.05, * p<0.1.

### Country factors and interactions (Table 3) — level effects
- Presence of capital controls is associated with a lower level of cross-border banking flows:
  - Capital controls coefficients reported: -0.0496** -0.0384* -0.0352** -0.0226 (standard errors in parentheses).
- Low institutional quality associated with lower cross-border flows:
  - Institution quality coefficients reported: -2.104*** -2.196*** -1.734*** -1.809*** (standard errors in parentheses). (Note: high values indicate lower institutional quality.)
- More restrictions on foreign bank entry reduce the level of cross-border change in claims on both banks and non-banks:
  - Limits on foreign banks coefficients reported: -1.128*** 1.883** -0.819*** 0.400 (standard errors in parentheses).

### Country factors and interactions (Table 3) — interaction with global liquidity (cyclicality)
- Exchange rate flexibility*Global_CB_Liquidity: -0.630*** -0.196 (standard errors in parentheses) — a more flexible exchange rate regime reduces the cyclicality of change in claims on banks.
- Capital controls*Global_CB_Liquidity: -0.0307 -0.0194** (standard errors in parentheses) — stricter capital controls reduce the cyclicality of change in claims on non-banks.
- Capital stringency*Global_CB_Liquidity: -0.207*** -0.0484 (standard errors in parentheses) — greater capital stringency reduces cyclicality of claims on banks.
- Supervisory power*Global_CB_Liquidity: -0.0936** -0.00438 (standard errors in parentheses) — greater supervision power reduces cyclicality of claims on banks.
- Limits on foreign banks*Global_CB_Liquidity: -3.360*** -1.361*** (standard errors in parentheses) — more restrictions on foreign bank entry lower the cyclical role of the global liquidity factor.

### Panel regression design notes
- The tables report estimates of panel regressions with country fixed effects and clustered standard errors at the borrower country level.
- Dependent variables: change in cross-border claims on banks and non-banks (log changes in BIS locational cross-border claims, in percent).
- Control variables included (but not always reported in displayed columns): lag GDP growth, lag CPI inflation, change in interest rate differentials, Global liquidity (PCA_Quantity), and other listed macro-financial controls (VIX, TED spread, global bank leverage, credit conditions in the US, real federal funds rate, slope of yield curve, G4 Countries M2).

### Asset management and international allocation (Annex IV) — size and trends
- Selected advanced economies — Assets under management of mutual funds and institutional investors (in trillions of U.S. dollars, unless otherwise indicated) by sample years:
  - Total assets: 22 35 53 68 65 70 76
  - Total in percent of GDP: 96 143 159 179 172 167 180
  - Mutual funds: 6 13 19 26 25 26 29
  - Share of open-end funds in total mutual fund assets (in percent): 94 97 96 97 97 97 97
  - Institutional investors: 16 23 34 41 40 44 47
  - Share of institutional investors in total assets (in percent): 72 64 64 61 61 63 61
- Selected emerging market and other economies:
  - Total assets: ......2.3 4.4 4.8 6.4 ...
  - Total in percent of GDP: ......32 36 37 36 ...
  - Mutual funds: ......0.8 1.9 1.9 2.3 ...
  - Institutional investors: ......1.5 2.5 2.9 4.1 ...
  - Share of institutional investors in total assets (in percent): ......65 59 60 62 ...
- Observations:
  - Growth of the asset management industry in both advanced and emerging economies has increased its role in global allocation of liquidity.
  - Concentration in the asset management industry has increased since the global financial crisis.
  - The official sector (sovereign wealth funds and reserves) has grown in importance in global asset allocation.

### Portfolio flows channel and data (Annex V) — methodology and indicators
- EPFR Global mutual fund country flows used as proxy for transmission of liquidity through international portfolio flows:
  - Monthly data on country level portfolio flows available from 1996 for equity flows and from 2004 for bond flows.
  - EPFR coverage expands over time; data are normalized by assets under management (AUM).
- Proxies for liquidity indicators sourced from multiple centers (United States, United Kingdom, Euro Area, Japan) and standardized so that an increase indicates more ample liquidity.
- Liquidity variables organized into 5 bins: (i) price and term premia; (ii) quantity; (iii) volatility; (iv) credit; (v) monetary policy.
- Global and country level controls include S&P500 excess returns and International Country Risk Guide (ICRG) composite risk rating.

### Representative list of liquidity indicators used (selection from Table 1)
- Price / term premia indicators: On-the-run vs Off-the-run 10-year Treasury liquidity premium; 10-year/2-year Treasury yields spread; 2-year/3-month Treasury yields spread; 10-year/3-month Treasury yields spread; 3-month TED spread; VIX; MOVE; SMOVE.
- Quantity indicators: Change in Repo Market Volume; Commercial Paper Outstanding; Fixed Income Open Interest; Equities Open Interest; New US Corporate Debt Issuance; Broker-Dealer Leverage; Change in M2; Growth of M0.
- Volatility indicators: CBOE S&P 500 Volatility Index (VIX); 1-month and 3-month Merrill Lynch swaption volatility expectations (MOVE/SMOVE).
- Credit indicators: Net share issuance and net bond issuance across jurisdictions; new issuance measures by region (US, UK, EA, Japan).
- Monetary indicators: Federal Funds Rate; Shadow policy rates (Wu and Xia, 2013); changes in regional M2 or M4; deposit facility rates; ECB/Bank of Japan policy rates.

*International Monetary Fund — Global Liquidity: Issues for Surveillance (selected excerpts from Annex III, Annex IV, and Annex V tables and text contained in the provided content).*

### 2.      Empirical Specification

### 2.      Empirical Specification

### Empirical strategy
- Baseline model: panel with country-level fixed-effects and robust standard errors for each liquidity variable.
- Regression specification (as in source): 
  - where f_it is the bond or equity net flow to country i at month t, L_t is a measure of global liquidity (taken from Table 1), r_t^e is the S&P500’s excess return over the US short term interest rate, and ICRG_{it-1} is the Composite Risk Rating for country i at time t.
- Models run separately for EM and AE.
- All variables are standardized so coefficients are comparable.
- Variables are ranked by significance, coefficient size, and R^2.
- Out-of-sample estimates: 3-month ahead mean square prediction error, with a rolling window of at least 3 years.
- Robustness: results robust to alternate specifications (such as lags) and subsamples, including subsamples corresponding to “first phase” (2003 to August 2008) and “second phase” (September 2008 onwards).
- Final step: use top-ranked variable in each bin and perform a “horse race” to test how much of bond and equity flows can be explained by best-performing liquidity variables.

### Results — high-level findings
- Local credit conditions and global portfolio investment flows share several global liquidity drivers.
- Most global liquidity variables highly correlated with commonality in credit conditions are also significantly correlated with portfolio investment flows to EM and AE.
- Some variables only moderately or weakly correlated with local credit conditions (e.g., TED spread moderately; VIX weakly) are nevertheless important drivers of portfolio flows.
- Monetary policy conditions in the Euro area and Japan are highly correlated with portfolio flows.

### Key coefficient estimates (selected, exact values preserved)
- TED spread (3-month TED spread (LIBOR-Treasury)) coefficients:
  - EM bond: 0.498***
  - AE bond: 0.512***
  - EM equity: 0.094***
  - AE equity: 0.157***
- US Growth in Commercial Bank Loans coefficients:
  - EM bond: 0.355***
  - AE bond: 0.380***
  - EM equity: -0.003
  - AE equity: -0.196***
- Japanese 10-year/3-months sovereign yield spread coefficients:
  - EM bond: 0.145***
  - AE bond: 0.213***
  - EM equity: 0.205***
  - AE equity: 0.260***
- VIX (CBOE S&P 500 Volatility Index) coefficients:
  - EM bond: 0.328***
  - AE bond: 0.230***
  - EM equity: 0.198***
  - AE equity: 0.243***
- Other illustrative coefficients from Table 2 (selected):
  - USA 10 year Off-the-run Liquidity Premium: EM bond 0.230***; AE bond 0.151***; EM equity 0.087***; AE equity 0.100***
  - US Broker Dealer Leverage: EM bond -0.061***; AE bond -0.247***; EM equity 0.021*; AE equity 0.119***
  - EA Deposit facility rate: EM bond 0.236***; AE bond 0.398***; EM equity 0.125***; AE equity 0.098***
- R^2 and explanatory power (Table 3 / summary):
  - Bond flows:
    - EM: R^2 0.467
    - AE: R^2 0.465
  - Equity flows:
    - EM: R^2 0.233
    - AE: R^2 0.239

### Interpretation of results
- Association between global liquidity and portfolio investment is stronger for bond flows than for equity flows:
  - “Nearly a third” of variation in bond flows is explained by liquidity variables; for equity flows the figure is less than a fifth.
- For bond flows:
  - Highest ranked price variable: TED spread.
  - Highest ranked credit variable: growth in US commercial bank loans.
- For equity flows:
  - Price and volatility variables dominate: Japanese 10-year/3-months sovereign yield spread and the VIX are more important.

### Annex VI — Off-Shore Debt Issuance: recent patterns (selected findings)
- Favorable funding market conditions allowed more financial institutions and corporates to tap international debt markets; emerging market issuance boomed and risk premiums tightened.
- Advanced economies remain main users of offshore debt, but EM presence has grown; AEs’ reliance on offshore debt declined post-crisis while EM share increased.
- The increase in reliance on offshore debt markets has been largest for China, with Brazil and Russia recording sizable increases in issuance.
- Issuance out of offshore financial centers stabilized and even increased overall since 2010.
- Sectoral composition (Dealogic / BIS stock as of 2013q3):
  - China: by sector of parent company (Dealogic stock as of 2013q3, 155 billions of USD)
    - NFC 38%
    - Government 60%
    - Banks 0%
    - OFI 2%
  - China: by sector of issuing company (BIS stock as of 2013q3, 240 billions of USD)
    - NFC 9%
    - Government 6%
    - Banks 26%
    - OFI 59%
  - Brazil: by sector of parent company (Dealogic stock as of 2013q3, 248 billlions of USD)
    - NFC 15%
    - Government 18%
    - Banks 30%
    - OFI 37%
- Offshore financing for operations at home is sizable especially for China and Brazil; balance of payments statistics broadly confirm funds raised by offshore subsidiaries returned to parent companies as reverse investment (i.e., FDI).
- Note: unless intercompany lending is included, currency mismatch risk may not be fully captured in external debt metrics.

### Annex VII — Banks’ Noncore Liabilities as indicators of financial sector vulnerability (selected)
- Background / indicator construction:
  - Two noncore indicator approaches (Hahm, Shin & Shin 2012):
    - Noncore 1: sum of banks’ liability to the foreign sector + (M3 – M2).
    - Noncore 2: sum of banks’ liabilities to the foreign sector + banks’ liabilities to the non-banking financial sector.
  - Hahm et al (2012) used 2000–2010, 202 countries (data limitations reduced to ~40 for most estimations).
  - Crisis definitions in Hahm et al:
    - Currency: depreciation > 25% in one year and at least 10 percentage points more than previous year.
    - Credit: sharp increases in money market rates reaching top 3 percent tail of pooled in-sample distribution.
    - Stock market crisis: stock index changes fall to bottom 3 percent tail of pooled in-sample distribution.
- Update performed by staff:
  - Noncore indicators used: noncore1 and noncore2-to-GDP; noncore1 and noncore2-to-M2.
  - Timeframe expanded: 1980 to the first half of 2013.
  - Sample: 43 countries (large advanced and emerging market countries).
  - Control variables added: policy target rates, domestic inflation, 3-month-to-10-year sovereign bond yield spread, general government primary balance as share of GDP, Chinn-Ito index of financial openness, VIX index, and US noncore indicators (as proxy for global liquidity).
  - All independent variables lagged by two quarters.
  - Primary estimation: panel probit model with random effects; robustness check: conditional logit with fixed effects (qualitatively similar results).
- Main findings:
  - Noncore indicators are significant indicators of financial vulnerability; coefficients for noncore liquidity ratios are positive and most are significant at the 1 percent and 5 percent level (except for credit crisis).
  - Noncore indicators’ predictive power is consistent across banking, stock market, and currency crises when considered alone.
  - No qualitative difference between level ratios (noncore-to-GDP) and funding-structure ratios (noncore-to-M2) in predictive power.
  - For banking crises:
    - Noncore indicators remain significant even after controlling for country-specific and global factors.
  - For currency and stock market crises:
    - Noncore ratios’ predictive power diminishes or disappears once country-specific macro variables and global risk appetite are controlled for — implying weaker performance for non-banking-related crises.
  - Inclusion of US noncore indicators (proxy for global liquidity) had mixed effects:
    - For banking crises (Laeven & Valencia database), inclusion of global liquidity boosted model goodness-of-fit from 60 percent to 90 percent, while removing statistical significance from local liquidity indicators.
    - For other crisis types, inclusion of US liquidity did not offer added benefits.
  - Risk sentiment (VIX) was significant for banking and currency crises (mostly at 1 percent), but not a significant predictor of stock market crises.

*Italic source: IMF — _031114 - 2.      Empirical Specification (excerpt) — Global Liquidity: Issues for Surveillance*

### Appendix 1.  Countries, Indicator and Crisis Availability

### Appendix 1.  Countries, Indicator and Crisis Availability

### Country-level crisis and indicator availability
- Table lists country rows with indicator availability marked by "x" across columns: Noncore 1, Noncore 2, Banking Crisis 1, Currency Crisis 2, Credit Crisis 2, Stock Mkt Crisis 2.
- Example country rows as presented (preserve exact strings):
  - Argentinaxxoooo
  - Australiaxxo
  - Austriaxxo
  - Belgiumxxoo
  - Brazilxxoooo
  - Canadaxoo
  - Chinaxo
  - Cyprusx
  - Czech Republicxoo
  - Denmarkxoo
  - Estoniao
  - Euro Areax
  - Finlandxxo
  - Francexxoo
  - Germanyxxo
  - Greecexxo
  - Hong Kong SARx
  - Indiaxoo
  - Indonesiaooo
  - Irelandxxo
  - Italyxxoo
  - Japanxo
  - Koreax
  - Luxembourgxxo
  - Malaysiaxo
  - Maltao
  - Mexicoxoo
  - Netherlandsxxoo
  - New Zealandxo
  - Portugalxxo
  - Russiaxo
  - Saudi Arabiax
  - Singaporex
  - Slovak Republico
  - Sloveni a o
  - South Africaxxoo
  - Spainxo
  - Swedenxxoo
  - Switzerlandxo
  - Thailandxo
  - Turkeyxoooo
  - United Kingdomxooo
  - United Statesxoo
- The table structure suggests variability in availability of noncore liquidity indicators and crisis event classification across countries, implying idiosyncratic causes and propagation of financial crises.

### Conclusion (from source)
- This updated empirical analysis of the Hahm, Shin & Shin (2012) exposition of noncore liquidity indicators generally affirmed the usefulness of these indicators as a predictor of financial crises; the results are particularly strong for banking crises.
- While the weak performance of these indicators in currency, credit and stock market crises suggest that their broad use may be limited, it nonetheless suggests that noncore banking liability indicators can be useful complements to other liquidity indicators in assessing financial sector vulnerability.

### Annex VIII. Network Representation of Cross-Border Claims
- Objective:
  - Explore current network structure of cross-border claims on portfolio debt and equities and structure changes compared to the pre-crisis period.
- Network representation:
  - Network structure is characterized by the relative size of bilateral claims (relative thickness of links) and centrality of individual nodes (size of nodes).
  - Relative thickness of links is calculated as the geometric average of the shares of each country pair’s bilateral claims in each country’s total claims, using notation Claim AB→ to represent country A’s claim on country B.
  - Claim AB→ represents country A’s claim on country B.
  - N
    i
    iB
    Claim  AI
    =
    (expression in source) represents country A’s total claim (as presented in the source).
  - Node size is a function of the number of links and the size of the links that travel to the node; a larger node represents higher centrality in a module.
- Interpretation and key findings:
  - The United States, the United Kingdom, Euro area economies (e.g. France, Germany, but also Luxembourg), and Japan remain central in the current bond and equity portfolio networks.
  - Japan, Singapore and Hong Kong serve as the financial centers connecting emerging Asia to the United States and Europe.
  - Nordic (Sweden, Finland, Denmark and Norway) and Central European (e.g. Poland, Hungary, and Czech Republic) economies form strong linkages amongst themselves, and connect to the core Europe through Austria.
  - Latin American economies have direct portfolio exposure with the United States, but bilateral claims within the region are relatively small.
  - The network structure of cross-border debt and equity holdings is little changed between 2006Q4 and 2012Q4, notwithstanding the overall decline in magnitude of cross-border exposures post-crisis.
  - Europe’s portfolio linkages with the United States, largely through the United Kingdom, Germany and Luxembourg, have weakened since the crisis.
  - China’s cross-border exposure with financial centers, e.g. Singapore and Hong Kong has increased significantly for both debt and equity securities.
  - The increased bilateral portfolio debt exposure between China and Hong Kong in part reflects the recent development of offshore debt issuance.
- Visualization notes:
  - For readability of the network chart, only the top 10 percent strongest links are depicted.
  - Portfolio Debt Bilateral Claims: 2006 Q4
  - Portfolio Debt Bilateral Claims: 2012 Q4
  - Portfolio Equity Bilateral Claims: 2006 Q4
  - Portfolio Equity Bilateral Claims: 2012 Q4
- Data source and tools (as stated in the source):
  - Source: CPIS, 2012 Q4.
  - Network map prepared by staff using the software from Rosvall and Bergstrom (2008).

*Prepared by Yuanyan Sophia Zhang (SPR) — content extracted from the supplied source.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/np/pp/eng/2014/_031114.pdf_