## _wp12246

## Source details

**Canonical URL:** [_wp12246](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12246.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12246.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12246.pdf.json)

---

### Inventory of exhibits
- Tables:
  - 1. Unit Root Tests of Liquidity Price and Quantity Indicators
  - 2. Estimation of Linear Trends
  - 3. Impact of Funding Shocks on Growth: Benchmark Model
  - 4. Impact of Funding Shocks on Growth: Separate Time Periods
  - 5. Impact of Funding Shocks on Growth: G4-Specific Impacts, Q1 1999–Q1 2011
- Figures:
  - 1–15 (including Total G4 Liquidity, National Measures, Supply and Demand Shocks, and Impact of Core/Noncore Shocks on Real GDP)
- Boxes:
  - 1. Estimation of the Noncore Liquidity Price Index
  - 2. Indentifying Demand and Supply Shocks
  - 3. A Panel Regression Approach to Assessing the Real Impact of Global Liquidity
  - 4. A VAR Approach to Assessing the Real Impact of Global Liquidity
- Appendix:
  - Measuring G4 Core and Noncore Liabilities

### Introduction and motivation
- Recent financial crises in the U.S. and Europe highlighted liquidity’s impact on economic and financial stability.
- Liquidity defined as the amount of funding readily available to finance domestic and cross-border asset purchases; reflects ability and willingness to transact and market capacity to absorb temporary fluctuations without undue price dislocations.
- Measurement challenges:
  - liquidity is largely endogenous and highly cyclical;
  - central bank injection of base money is important but global liquidity also driven by growth differentials, financial innovation, and market participants’ risk appetite;
  - collateralized market-based borrowing introduces endogeneity as funding expands or contracts with collateral valuations.
- Combining price and quantity measures:
  - quantity indicators reflect size of risk exposure and tend to be slow-moving;
  - price indicators are coincident and spike when crises are underway;
  - combined analysis helps disentangle supply (persistent increases in liquidity supply → growing quantity and falling prices) from demand (higher demand → increases in both price and quantity).
- Caveats:
  - no theoretical framework for an optimal level of global liquidity;
  - rapid transformation of financial markets and data shortcomings;
  - policy conclusions should rely on thorough analysis and further research.

### Core and noncore liquidity: definitions and aggregates
- Core liquidity:
  - total resident deposits in commercial banks and other depositary corporations;
  - broadly similar to traditional broad money aggregates (particularly M3);
  - excludes inter-bank deposits (deposits of other financial institutions with commercial banks and other depositary corporations).
- Noncore liquidity:
  - total nonresident deposits in commercial banks and other depositary corporations (cross-border deposits considered noncore) and loans and securities (other than shares) of commercial banks, nonbanks and other financial intermediaries;
  - includes liabilities across financial institutions and thus captures potential double counting from rehypothecation and other collateral reuse.
- Aggregate observations:
  - Global liquidity has more than doubled since 1999 in nominal terms.
  - Two thirds of the rise is attributable to noncore liquidity, particularly since 2004.
  - Noncore global liquidity experienced deviations from trend of around 15 percent of G4 GDP; deviations around 6 percent for core liquidity.

### Price measures and the Noncore Liquidity Price Index (NLPI)
- Price of core liquidity:
  - spread between domestic deposit rate for deposits with a maturity of up to one year and the 6-month interbank offered rate;
  - spreads standardized; global core price is weighted average with weights equal to country core liabilities as a ratio of global core liabilities.
- Noncore Liquidity Price Index (NLPI):
  - captures marginal cost of noncore funding using interest rate spreads, asset prices, credit volume, and lending condition surveys;
  - constructed with a dynamic factor model decomposing each indicator into a common and idiosyncratic component;
  - index in standard deviations from the average, with the financial crisis of 2008 at four standard deviations;
  - empirical behavior: the global NLPI jumped by more than four standard deviations in 2008; country-specific NLPIs also peaked in 2008; Japan exhibited smaller fluctuations.

### Funding-based approach and supply–demand framework
- Conceptual shift: liability-side/funding-based approach to capture bank and nonbank funding, including wholesale and collateral-based financing.
- Core vs noncore roles:
  - Core: nonfinancial sector supplies; retail banks and money market funds demand.
  - Noncore: financial institutions can be both suppliers and demanders; rehypothecation implies potential double counting.
- Price–quantity sign interpretation:
  - Demand shock: prices and quantities move in the same direction.
  - Supply shock: prices and quantities move in opposite directions.
- Policy-relevant dynamics:
  - Persistent increasing supply (rising quantities, falling prices) → increased risk taking/leverage → possible macroprudential or regulatory measures.
  - Growing demand (rising prices) may restrain overinvestment but signals deteriorating fundamentals and requires monitoring.

### Identification and historical dynamics (VAR with sign restrictions)
- Data and decomposition:
  - Quantity indicators contain a unit root (per ADF and PP tests) and are decomposed into trend and cycle; a linear deterministic trend is removed in practice.
  - Trend growth of liabilities (as a ratio to GDP): Noncore liabilities: 1 percentage point per quarter; Core liabilities: half a point per quarter.
- VAR identification:
  - use of unrestricted VAR x_t = B(L) x_{t-1} + ε_t and structural identification via Cholesky decomposition and rotations with sign restrictions;
  - “median targeting” selects median impulse response among acceptable rotations;
  - minimum number of models between 3,000 or until difference between median responses of two sets of models is reasonably small.
- Historical decomposition — Core liquidity (three periods):
  - 1999–2001: positive supply shocks outweighed by negative demand shocks; core liquidity (ratio to GDP) slowly reduced.
  - Precrisis up to 2008: negative supply shocks brought core liquidity below trend as investors sought higher-yield instruments.
  - Global financial crisis: positive demand and supply shocks kept core liquidity above trend (central bank injections).
- Historical decomposition — Noncore liquidity (four periods):
  - Early period: positive supply shocks trended noncore upward but below trend.
  - 2005–07: positive supply shocks drove rapid growth in noncore liquidity (inside money creation, leverage).
  - 2008 crisis: falling supply and rising demand; quantity stable but price increased sharply.
  - 2011 onwards: negative supply shocks depressed price and quantity (deleveraging in shadow banking).
- Country-level differences (G4):
  - US and euro area: core liquidity dynamics largely demand-driven; US core demand fell precrisis then rose sharply in 2008.
  - UK: precrisis core dynamics mainly supply-driven.
  - Japan: falling nominal GDP produced a unique path; core liquidity to GDP rose sharply at crisis onset, largely supply-driven.
  - Noncore liabilities: US rise largely supply-driven; UK and euro area showed stronger demand for noncore liabilities.
- External liabilities to BIS reporting banks:
  - G4 noncore global liabilities behave similarly to G4 external liabilities to BIS reporting banks.
  - Non-G4 countries: external liabilities larger as ratio to GDP; demand-driven trends with dramatic fall in 2008 followed by rapid recovery.

### Real impact of global liquidity on growth (panel and VAR evidence)
- Data and methods:
  - Unbalanced panel Q1 1999 to Q1 2011 covering G4 and other advanced and emerging economies.
  - Two approaches: panel regression (Arellano-Bond two-step GMM) and country-specific VAR impulse responses.
  - Panel explanatory variables include: R (policy rate), P (headline inflation), CS (core supply shock), CD (core demand shock), NCS (noncore supply shock), NCD (noncore demand shock). Interaction dummies: G4, crisis (Q4 2007–Q2 2009), post-crisis.
- Key quantitative findings (preserve numeric magnitudes and signs exactly as reported):
  - Demand shocks to liquidity have stronger effects on real GDP than supply shocks.
    - Demand shock to noncore liquidity: average impact of -0.98 percentage points of GDP after eight quarters for G4; -0.90 percentage points of GDP for non-G4 countries.
  - Duration of effects:
    - Demand-driven shocks: long-lasting effects; supply-driven shocks tend to fade by eight quarters.
    - Example: supply shock to noncore reduces to 0.18 percentage points impact after eight quarters (compared to -0.98 for demand).
  - Core liquidity shocks:
    - A positive demand (supply) shock to core liquidity lowers real GDP by 0.5 (0.3) percentage points after eight quarters.
  - Noncore liquidity shocks:
    - Positive supply shocks to noncore generally have positive effects on GDP growth (pro-cyclical).
    - Demand shocks to noncore are counter-cyclical and have large negative spillovers.
  - Temporal variation:
    - Precrisis: strongly positive impact of noncore supply shocks drove growth.
    - Crisis: liquidity impact on GDP negative, largely due to core liquidity shocks.
    - Post-crisis: impact turned positive, driven by demand shocks to core liquidity and, to a lesser extent, noncore liquidity—consistent with regulatory shifts (e.g., Basel III liquidity requirements, Dodd-Frank Act, Capital Requirement Directive IV).
  - Cross-country sensitivity:
    - No statistically significant cross-country variations in sensitivity to G4 funding shocks were found.

### Key empirical tables and selected numeric statistics (preserved exactly)
- Table 1. Unit Root Tests of Liquidity Price and Quantity Indicators, p-values
  - Core liquidity price index: ADF 0.1610, PP 0.4556
  - Noncore liquidity price index: ADF 0.4546, PP 0.6538
  - G4 core liquidity, ratio to GDP: ADF 0.9143, PP 0.8834
  - G4 noncore liquidity, ratio to GDP: ADF 0.5461, PP 0.9991
  - G4 core liquidity, trillion dollars: ADF 0.0465, PP 0.4154
  - G4 noncore liquidity, trillion dollars: ADF 0.8765, PP 0.9305
- Table 2. Estimation of Linear Trends
  - Core Liquidity, Ratio to GDP: Constant 0.8379***; Time 0.0054***; Adjusted R-squared 0.8201; Durbin Watson statistic 0.1256
  - Noncore Liquidity, Ratio to GDP: Constant 1.0516***; Time 0.0103***; Adjusted R-squared 0.8175; Durbin Watson statistic 0.1210
  - Notes: *** denotes significance at 1 percent.
- Table 3. Impact of Funding Shocks on Growth: Benchmark Model, Q1 1999–Q1 2011 (selected coefficients)
  - Real GDP growth rate t-1 coefficients: 0.318***, 0.304***, 0.267***, 0.413***, 0.450***, 0.437***, 0.347***, 0.392***, 0.364*** (z-statistics in parentheses)
  - Policy rate t-1 coefficients include -0.212*, -0.134, -0.158, -0.177, -0.118, -0.114, -0.241**, -0.040, -0.197** (z-statistics in parentheses)
  - Headline inflation t coefficients include -0.175***, -0.112, -0.117, -0.158*, -0.136*, -0.131*, -0.069, -0.158***, -0.059 (z-statistics in parentheses)
  - Core liquidity supply shock during crisis t-1: -1.094***, -0.697**, -0.941*, -0.476 (z-statistics shown)
  - Core liquidity demand shock during crisis t-1: -1.644***, -1.525***, -1.314***, -1.394*** (z-statistics shown)
  - Noncore liquidity supply shock t-1: 0.205***, 0.198, -0.024, 0.214 (z-statistics shown)
  - Noncore liquidity demand shock during crisis t-1: -0.836***, -0.594*, -0.110, -0.024 (z-statistics shown)
  - Constants reported (e.g., 5.041***, 4.054***, 4.661***, ...), Number of groups 33, Number of observations 1,192
  - Note: ***,**, * denote significance at 1%, 5% and 10%, respectively. Two-step using Windmeijer standard errors.
- Table 4. Impact of Funding Shocks on Growth: Separate Time Periods, Q1 1999–Q1 2011
  - Real GDP growth rate t-1: 0.305***, 0.302***, 0.296***, 0.420***, 0.452***, 0.435***, 0.341***, 0.392***, 0.328***
  - Core liquidity supply shock during crisis t-1: -1.022***, -0.545**, -0.587***, -0.677* (selected entries)
  - Constants and sample sizes as reported; Number of groups 33; Number of observations 1,192
  - Note: ***,**, * denote significance at 1%, 5% and 10%, respectively. Two-step using Windmeijer standard errors.
- Table 5. Impact of Funding Shocks on Growth: G4-Specific Impacts, Q1 1999–Q1 2011
  - Real GDP growth rate t-1: 0.361***, 0.358***, ... across specifications
  - G4 core supply shock t-1: -0.449***, -0.462***, -0.487***, -0.466***
  - G4 core demand shock t-1: -0.303***, -0.311***, -0.293***, -0.329***
  - G4 noncore supply shock t-1: 0.262***, 0.250***, 0.203**, 0.217***
  - G4 noncore demand shock t-1: -0.543***, -0.546***, -0.561***, -0.566***
  - Constants and sample sizes as reported; Number of groups 33.000; Number of observations 1,192
  - Note: ***,**, * denote significance at 1%, 5% and 10%, respectively. Two-step using Windmeijer standard errors.

### Methodological boxes — NLPI, VAR, Panel, and bootstrap CI construction
- Box 1 (NLPI, Dynamic Factor Model):
  - y_t(i) = λ_i F_t + ε_t(i), ε_t(i) ~ N(0, Ω) with Ω diagonal.
  - F_t = Σ_{j=1}^k β_j F_{t-j} + u_t, u_t ~ N(0,1); lag length k chosen by Swartz-Bayesian information criteria.
  - State-space estimated via Kalman filter; weights determined optimally by Kalman recursion.
  - Mixed-frequency and uneven sample length handling; missing data back-cast per Giannone, Reichlin, and Small (2008).
  - Global NLPI constructed by pooling G4 indicators.
- Box 2 (Identifying demand and supply shocks):
  - Quantity indicators contain a unit root; de-trended price and quantity used in VAR.
  - VAR: x_t = B(L) x_{t-1} + ε_t; structural identification via P with rotations Q; sign restrictions select rotations matching theoretical signs.
  - “Median targeting” selects median impulse response among iterations; small sample limits time-invariant VAR.
  - Historical decomposition follows representation y_t = μ + Ψ_1 e_d^t + ... + Π_1 e_s^t + ...
- Box 3 (Panel regression: Arellano-Bond GMM):
  - Dependent y_{i,t} is year-on-year real growth.
  - Explanatory variables: lags of y, R, P, CS, CD, NCS, NCD; instrumenting addresses endogeneity and autocorrelation; fixed effects handled.
  - Interaction dummies: G4, crisis (Q4 2007–Q2 2009), post-crisis.
- Box 4 (VAR approach to real impact):
  - Uses VAR as in Box 2 to compute cumulative impulse response functions over two, four and eight quarter horizons.
  - IRF_{(i,j,h)} = q' A^h s with companion matrix Γ (notation per VAR companion form).
- Section IV (Bootstrap bias-correction and confidence intervals):
  - Bootstrap procedure following Kilian (1998):
    - collect estimates ߚ̂ = vec( ̂B( ̂Ψ ) );
    - resample residuals and simulate to obtain a thousand additional samples and coefficient estimates; denote average by ߚ̄ and bias ߚΔ = ߚ̄ − ߚ̂;
    - construct bias-corrected estimate ߚ̃ with iterative adjustments if companion-matrix eigenvalues outside unit circle (steps described with adjustments using ߜ and Δ, and iteration until stationarity);
    - final bootstrap uses ߚ̃ to generate samples and obtain bias-corrected 90 percent confidence interval (5-th and 95-th percentile) and median estimate.
  - Figures 12–15 summarize confidence intervals of impacts of one standard deviation shocks to supply/demand of core and noncore liquidity on real GDP (Change in Level of Real GDP after 2, 4, and 8 quarters; dashed line denotes subgroup average; 5th and 95th percentiles and median plotted).
  - Notes: alternative approaches include Monte Carlo integration and Bayesian asymptotic intervals; small sample size is a key limitation.

### Stylized policy-relevant conclusions and surveillance implications
- Liability type and shock nature matter for macro outcomes.
- Noncore liquidity:
  - more endogenous and procyclical; expansion via leverage and low prices can build vulnerabilities;
  - negative shocks can cause sizeable GDP declines via deleveraging and tighter funding.
- Supply vs demand interpretation:
  - Prolonged low funding prices with rising quantities (positive supply shocks) raise systemic risk and may warrant macroprudential/regulatory restraint on leverage.
  - Demand-driven increases (higher prices) can be stabilizing by restraining overinvestment, but reflect worsening fundamentals and require monitoring.
- Surveillance recommendations:
  - Monitor both price and quantity measures; their interaction helps identify supply vs demand drivers.
  - Capture liquidity creation from traditional bank channels and volatile nonbank channels for financial stability surveillance.
- Research and measurement gaps:
  - Accurate price indicators are critical but difficult to establish.
  - More work needed on structural breaks, financial innovation, and properties of global liquidity.
- Policy caveat:
  - No universal “cookbook”; appropriate response depends on financial structure, openness, monetary autonomy, and shock origin.
  - Goodhart’s Law: indicators can lose forecasting value as actors change behavior; supply–demand interpretation remains informative.

*Source: _wp12246 (excerpted boxes, Chapter 1, Section IV, and References as provided).*

### References .............................................................................................................

### _wp12246 - References

### Inventory of Exhibits
- Tables listed:
  - 1. Unit Root Tests of Liquidity Price and Quantity Indicators
  - 2. Estimation of Linear Trends
  - 3. Impact of Funding Shocks on Growth: Benchmark Model
  - 4. Impact of Funding Shocks on Growth: Separate Time Periods
  - 5. Impact of Funding Shocks on Growth: G4-Specific Impacts, Q1 1999–Q1 2011
- Figures listed:
  - 1. Total G4 Liquidity in Trillion Dollars and As a Ratio to GDP
  - 2. National Measures of the Quantity of Liquidity, Ratio to National Nominal GDP
  - 3. National Measures of the Quantity of Liquidity, Trillion US dollars
  - 4. Total External Liabilities to BIS Reporting Banks
  - 5. Nominal GDP Growth Rates
  - 6. Supply and Demand Shocks, Quantity and Price of Core Global Liquidity
  - 7. Supply and Demand Shocks, Quantity and Price of Noncore Global Liquidity
  - 8. Supply and Demand Shocks to Liquidity: United States and Euro Area
  - 9. Country-Specific Supply and Demand Shocks: United Kingdom and Japan
  - 10. Supply and Demand Shocks: External Liabilities of G4 Economies
  - 11. Supply and Demand Shocks: External Liabilities of other countries to BIS Reporting Banks
  - 12. Impact of Core Demand Shock on Real GDP
  - 13. Impact of Core Supply Shock on Real GDP
  - 14. Impact of Noncore Demand Shock on Real GDP
  - 15. Impact of Noncore Supply Shock on Real GDP
- Boxes listed:
  - 1. Estimation of the Noncore Liquidity Price Index
  - 2. Indentifying Demand and Supply Shocks
  - 3. A Panel Regression Approach to Assessing the Real Impact of Global Liquidity
  - 4. A VAR Approach to Assessing the Real Impact of Global Liquidity
- Appendix:
  - Measuring G4 Core and Noncore Liabilities

### Key points from the Introduction and motivation
- Recent financial crises in the U.S. and Europe highlighted the impact of liquidity on economic and financial stability.
- Liquidity is described as the amount of funding readily available to finance domestic and cross-border asset purchases and reflects both ability and willingness to transact and market capacity to absorb temporary fluctuations without undue price dislocations.
- Measuring global liquidity is challenging because:
  - liquidity is largely endogenous and highly cyclical;
  - central bank injection of base money is important but global liquidity also driven by growth differentials, financial innovation, and market participants’ risk appetite;
  - collateralized market-based borrowing introduces endogeneity as funding expands or contracts with collateral valuations.
- Combining price and quantity measures provides richer analysis:
  - quantity indicators reflect size of risk exposure and tend to be slow-moving;
  - price indicators are coincident and spike when crises are underway;
  - combined analysis helps disentangle supply (persistent increases in liquidity supply → growing quantity and falling prices) from demand (higher demand → increases in both price and quantity).
- Caveats emphasized:
  - no theoretical framework for an optimal level of global liquidity;
  - rapid transformation of financial markets and data shortcomings (few countries compute flow of funds; cross-country reporting inconsistency);
  - policy conclusions should rely on thorough analysis and further research.

### Literature review: empirical and conceptual findings
- Evidence linking monetary aggregates to macro outcomes:
  - Sousa and Zaghini (2004): changes in global (excluding euro) liquidity explain a significant share of euro area price and output fluctuations.
  - Ruffer and Stracca (2006): positive shock to global excess liquidity (broad money to nominal GDP) leads to significant rise in domestic real output and price level for the euro area and Japan.
  - Global liquidity affects asset and commodity prices (Darius and Radde, 2010; Thomas, Muhleisen and Pant, 2010; Psalida and Tao, 2011).
- Shadow banking and market-based funding:
  - Adrian, Ashcraft, Boesky and Poszar (2010): volume of credit intermediated by shadow banking has exceeded traditional banking since mid-1990s.
  - Adrian and Shin (2010): balance sheet expansion of investment banks proxies overall funding conditions in market-based systems.
- Cross-border liquidity linkages and funding risk:
  - Shin and Shin (2011): increase in noncore liabilities predicts currency appreciation and rising credit spreads in Korea.
  - McGuire and von Goetz (2009): European banks’ reliance on interbank borrowing and dollar funding exposed them to funding risk during the crisis.
  - Bruno and Shin (2011): fluctuating leverage of cross-border banks transmits global financial conditions.
- Leverage, collateral and pro-cyclicality:
  - Collateralized borrowing amplifies risk propagation (references: Bernanke and Gertler, 1989; Kiyotaki and Moore, 1997; Bernanke, Gertler and Gilchrist, 1999).
  - Jorda et al (2011): credit growth is the single best predictor of financial instability across 14 developed countries over 140 years.
  - Pro-cyclical margin requirements can create liquidity spirals (Geanakoplos, 2009; Brunnermeier and Sannikov, 2010).
  - Singh (2011): use and reuse of collateral lubricates the financial system; decline in such collateral likely reduces market liquidity.

### A new funding-based approach to measuring global liquidity
- Conceptual shift:
  - Move beyond traditional monetary aggregates toward a liability-side/funding-based approach to capture bank and nonbank funding, including wholesale and collateral-based financing.
  - “Core” liabilities: relatively stable funding (example: retail deposits of the household sector).
  - “Noncore” liabilities: funding that co-varies with the financial cycle (wholesale and collateral-based market funding, shadow banking).
- Rationale:
  - Funding markets represent the balance-sheet counterpart to intermediated lending; measured risks (e.g., banks' Value-at-Risk) drive expansion/contraction of intermediaries.
  - Financial institutions can be both suppliers and consumers of noncore liabilities; financial innovation expands collateral and thus potential funding.
  - Liability aggregates convey information on degree of risk-taking and vulnerability to reversals.

### A. Quantity measures (operational details and data scope)
- Definition:
  - Quantity of global liquidity is defined as the sum of financial sector liabilities of the euro area, Japan, the United Kingdom, and the United States (the “G4” economies).
- Aggregation and normalization:
  - Global liquidity is computed by aggregating individual liquidity series across the G4 economies.
  - G4 aggregates are expressed in U.S. dollars and normalized by the (U.S. dollar denominated) nominal GDP of the four countries.
  - In practice, a linear deterministic trend is removed from the quantity indicators.
- Data coverage and inclusions:
  - Financial liabilities are taken from aggregate financial sector balance sheets captured in flow of funds accounts of the respective economies.
  - Included: bank and nonbank intermediaries raising funding in capital markets on a collateralized basis; money market funds; liabilities of insurance companies and pension funds.
- Justification for G4 focus and USD denomination:
  - G4 accounted for between 82 and 92 percent of total BIS external claims since 1995.
  - US dollars denomination is used because much of financial sector liabilities are denominated in US dollars.
- Empirical notes:
  - For Japan, liabilities (in trillions of dollars) are less volatile than when measured as a ratio to nominal GDP, because Japanese nominal GDP declined several times over the sample period (Figure 5).
  - Prime money market funds in the United States have held roughly 70 to 80 percent of their assets in the liabilities of the banking sector in recent years.

*Source: _wp12246 - References, pages and sections as provided in the source content.*

### Chapter 1).

### _wp12246 - Chapter 1)

### Definitions: Core and Noncore Global Liquidity
- Core liquidity:
  - Defined as total resident deposits in commercial banks and other depositary corporations.
  - Broadly similar to traditional broad money aggregates, particularly M3.
  - Excludes inter-bank deposits (deposits of other financial institutions with commercial banks and other depositary corporations) because these do not typically represent a source of “liquidity” for the nonfinancial private sector—i.e., they do not create leverage.
- Noncore liquidity:
  - Defined as total nonresident deposits in commercial banks and other depositary corporations (hence cross-border deposits are considered noncore) and loans and securities (other than shares) of commercial banks, nonbanks and other financial intermediaries.
  - Includes liabilities across financial institutions and thus captures potential double counting from rehypothecation and other collateral reuse, which helps assess gross leverage and funding available.
- Key aggregate observations:
  - Global liquidity has more than doubled since 1999 in nominal terms.
  - Two thirds of the rise is attributable to noncore liquidity, particularly since 2004.
  - Noncore global liquidity experienced deviations from trend of around 15 percent of G4 GDP, while these deviations were around 6 percent in the case of core liquidity.

### Price Measures for Liquidity
- Price of core liquidity:
  - Defined as the spread between domestic deposit rate for deposits with a maturity of up to one year and the 6-month interbank offered rate.
  - Spreads are standardized to control for cross-border differences in deposit coverage.
  - Global core price is the weighted average of individual countries’ standardized spreads, with weights equal to country core liabilities as a ratio of global core liabilities.
- Noncore Liquidity Price Index (NLPI):
  - Captures marginal cost of noncore funding using interest rate spreads, asset prices, credit volume, and lending condition surveys.
  - Constructed with a dynamic factor model decomposing each indicator into a common and idiosyncratic component.
  - The index is in standard deviations from the average, with the financial crisis of 2008 at four standard deviations.
  - Empirical behavior:
    - The global NLPI jumped by more than four standard deviations in 2008.
    - Country-specific NLPIs also peaked in 2008; Japan exhibited smaller fluctuations in price and quantity of noncore liquidity relative to other G4 economies.

### Liquidity Supply and Demand Framework
- Conceptual assignment:
  - Core liquidity: nonfinancial sector supplies; retail banks and money market funds demand.
    - Positive supply shock example: household shift to bank deposits (“flight-to-safety”).
    - Positive demand shock example: banks forced to raise more stable (and expensive) funding due to regulation or market stress.
  - Noncore liquidity: financial institutions can be both suppliers and demanders; rehypothecation implies potential double counting.
    - Positive supply shock example: “inside” money creation as global banks raise wholesale funding to leverage balance sheets amid low liquidity prices.
    - Positive demand shock example: institutions with high rollover needs bidding up noncore funding price during deteriorating fundamentals.
- Price–quantity sign interpretation:
  - Demand shock: prices and quantities move in the same direction (parallel shift of negatively sloped demand curve).
  - Supply shock: prices and quantities move in opposite directions (parallel shift of positively sloped supply curve).
- Policy-relevant dynamics:
  - Persistent increasing supply (rising quantities, falling prices) is associated with increased risk taking and leverage and may require macro-financial measures.
  - Growing demand for liquidity (rising prices) may reduce overinvestment risks but still needs monitoring if driven by expectations of future productivity.

### Assessing Drivers of Global Liquidity (Identification & Historical Dynamics)
- Identification:
  - Use of a VAR model with sign restrictions to separate demand and supply shocks.
  - Demand shocks: contemporaneous increase in both price and quantity.
  - Supply shocks: increase in quantity and decrease in price.
  - Supply of liquidity equals supply of funding (reverses conventional money terminology).
- Historical decomposition — core liquidity (three periods):
  - 1999–2001: positive supply shocks outweighed by negative demand shocks; core liquidity (ratio to GDP) slowly reduced.
  - Precrisis up to 2008: negative supply shocks brought core liquidity well below trend as investors sought higher-yield instruments.
  - Global financial crisis: positive demand and supply shocks kept core liquidity above trend (reflecting central bank injections).
- Historical decomposition — noncore liquidity (four periods):
  - Early period: positive supply shocks allowed noncore liquidity to trend upward but remained below trend.
  - 2005–07: positive supply shocks drove rapid growth in noncore liquidity consistent with “inside” money creation, leverage, and financial innovation.
  - 2008 crisis: falling supply and rising demand; quantity stable but price increased sharply.
  - 2011 onwards: negative supply shocks depressed both price and quantity of noncore liquidity (deleveraging in shadow banking).
- Country-level differences (G4):
  - United States and euro area: core liquidity dynamics largely demand-driven; US core demand fell precrisis then rose sharply in 2008.
  - United Kingdom: precrisis core dynamics driven mainly by supply factors.
  - Japan: unique path due to falling nominal GDP; core liquidity to GDP rose sharply at crisis onset, largely supply-driven.
  - Noncore liabilities:
    - US rise largely a supply phenomenon (pro-cyclical collateral-based “inside money” creation).
    - UK (and to lesser extent euro area) showed stronger demand for noncore liabilities, consistent with funding/reintermediation patterns across global banks.
- External liabilities to BIS banks:
  - G4-based noncore global liabilities behave similarly to G4 external liabilities to BIS reporting banks.
  - For non-G4 countries, external liabilities are relatively larger as a ratio of GDP and show demand-driven trends with a dramatic fall in 2008 followed by rapid recovery and reversion to trend.

### The Real Impact of Global Liquidity on Growth (Evidence)
- Data and methods:
  - Unbalanced panel from Q1 1999 to Q1 2011 covering G4 and a range of advanced and emerging economies.
  - Two approaches: panel regression (PR) and VAR country-specific analysis.
- Key quantitative findings:
  - Demand shocks to liquidity have stronger effects on real GDP than supply shocks.
    - Demand shock to noncore liquidity: average impact of -0.98 percentage points of GDP after eight quarters for G4; -0.90 percentage points of GDP for non-G4 countries.
  - Duration of effects:
    - Demand-driven shocks have long-lasting effects; supply-driven shocks tend to fade by eight quarters.
    - Example: supply shock to noncore reduces to 0.18 percentage points impact after eight quarters (compared to -0.98 for demand).
  - Core liquidity shocks:
    - Both demand and supply shocks to core liquidity are counter-cyclical.
    - A positive demand (supply) shock to core liquidity lowers real GDP by 0.5 (0.3) percentage points after eight quarters.
  - Noncore liquidity shocks:
    - Positive supply shocks to noncore generally have positive effects on GDP growth (pro-cyclical).
    - Demand shocks to noncore are counter-cyclical and have large negative spillovers.
  - Methodological nuances:
    - PR results sometimes differ from VAR; PR finds supply shocks to core liquidity positive on average (but reversed during crisis), while VAR shows counter-cyclical negative effects—reflecting differences between unconditional average instantaneous effects (PR) and dynamic/conditional responses (VAR).
  - Temporal variation:
    - Precrisis: strongly positive impact of noncore supply shocks drove growth.
    - Crisis: liquidity impact on GDP was negative, largely due to core liquidity shocks.
    - Post-crisis: impact of liquidity on GDP turned positive again, driven by demand shocks to core liquidity and, to a lesser extent, noncore liquidity—consistent with regulatory shifts encouraging core funding (e.g., Basel III liquidity requirements, Dodd-Frank Act, Capital Requirement Directive IV).
  - Cross-country sensitivity:
    - No statistically significant cross-country variations in sensitivity to G4 funding shocks were found.

### Stylized Policy-Relevant Conclusions
- Both the type of liability (core vs noncore) and the nature of shocks (supply vs demand) matter for macroeconomic outcomes.
- Noncore liquidity:
  - More endogenous and procyclical; expansion through leverage and low prices can build vulnerabilities.
  - Negative shocks can cause sizeable GDP declines via deleveraging and tightening funding conditions.
- Supply vs demand interpretation:
  - Prolonged periods of low funding prices with rising quantities (positive supply shocks) raise systemic risk and may warrant macroprudential or regulatory measures to restrain leverage.
  - Demand-driven increases (higher prices) can be stabilizing by restraining overinvestment; however, they reflect worsening fundamentals and require monitoring.
- Surveillance implications:
  - Monitoring both price and quantity measures is essential; their interaction helps identify supply vs demand drivers of liquidity expansions.
  - Capturing liquidity creation from traditional bank channels and from volatile nonbank channels is important for financial stability surveillance.
- Research and measurement gaps:
  - Accurate price indicators are critical for identifying supply and demand but can be difficult to establish.
  - More work is needed on structural breaks, financial innovation, and the properties of global liquidity.
- Policy caveat:
  - No universal “cookbook” for country responses; appropriate policy depends on financial structure, openness, monetary autonomy, and shock origin.
  - Goodhart’s Law cautions that indicators can lose forecasting value as actors change behavior; nonetheless, interpreting liquidity developments via supply and demand remains informative.

*Source: _wp12246 - Chapter 1).*

### Box 1. Estimation of the Noncore Liquidity Price Index

### Box 1. Estimation of the Noncore Liquidity Price Index

### Estimation framework (Dynamic Factor Model)
- Each standardized monthly financial indicator y_t(i) is decomposed as:
  - y_t(i) = λ_i F_t + ε_t(i), where ε_t(i) ~ N(0, Ω)
  - Ω is assumed diagonal (idiosyncratic components uncorrelated across indicators).
- The common factor F_t is the estimated noncore liquidity price index (NLPI) and follows an autoregressive (AR) k process:
  - F_t = Σ_{j=1}^k β_j F_{t-j} + u_t, where u_t ~ N(0,1)
- The lag length k is selected using the Swartz-Bayesian information criteria.
- The state-space system is estimated using the Kalman filter because the common factor is unobserved.

### Construction and data handling
- The estimated NLPI is a weighted average of the chosen financial indicators; weights are determined optimally by the Kalman filter recursion.
- The Kalman filter accommodates mixed frequencies and uneven sample lengths (e.g., bank lending surveys).
- Indicators include interest rate spreads, asset prices, risk appetite, and lending condition surveys.
- Missing data are completed using “back-casting” following the methodology of Giannone, Reichlin, and Small (2008).
- The global NLPI is constructed by pooling the G4 indicators.

### Key methodological points
- Use of Kalman filter for state-space estimation and optimal weighting.
- Mixed-frequency and uneven sample length handling via Kalman recursion.
- Back-casting for missing observations using Giannone, Reichlin, and Small (2008).

---

### Box 2. Identifying Demand and Supply Shocks

### Stationarity and trends
- Quantity indicators of global liquidity (both expressed as a ratio to GDP and in U.S. dollars) contain a unit root per augmented Dickey-Fuller (ADF) and Phillips-Perron tests.
- Quantity indicators are decomposed into trend and cycle components.
- Trend growth of liabilities (as a ratio to GDP):
  - Noncore liabilities: 1 percentage point per quarter.
  - Core liabilities: half a point per quarter.
- Deviations from trend are persistent; assuming a linear trend interprets stochastic variation as persistent supply and demand shocks.

### VAR identification with sign restrictions
- De-trended price and quantity indicators are regressed on lagged values using quarterly data from 1999Q1 to 2011Q1.
- Unrestricted VAR specification:
  - x_t = B(L) x_{t-1} + ε_t, where ε_t ~ N(0, Ω)
- Structural identification via Cholesky decomposition and rotations:
  - P such that P P' = Ω, with orthonormal matrix Q used to rotate P.
  - Structural residuals u_t have identity covariance after rotation.
- Sign restrictions are used to choose rotations whose impulse responses match theoretical signs.
- “Median targeting” approach is used to limit model choices by selecting the median impulse response among many iterations.
- Small sample size limits estimation to a time-invariant VAR.
- Minimum number of models is between 3,000 or the number such that the difference between median responses of two sets of models is reasonably small.

### Historical decomposition
- After selecting a model that matches sign restrictions, World historical decomposition is used to identify cumulative contributions of supply and demand shocks.
- Decomposition representation:
  - y_t = μ + Ψ_1 e_d^t + ... + Π_1 e_s^t + ...
- The approach follows procedures used in decomposing oil demand and supply factors.

---

### Box 3. A Panel Regression Approach to Assessing the Real Impact of Global Liquidity

### Model specification
- Dependent variable:
  - y_{i,t} is the real growth rate, year on year.
- Explanatory variables include:
  - R: country policy rate.
  - P: country headline inflation, year on year.
  - CS: supply shock on core liquidity.
  - CD: demand shock on core liquidity.
  - NCS: supply shock on noncore liquidity.
  - NCD: demand shock on noncore liquidity.
- Panel specification (parsimonious representation):
  - y_{i,t} = ... + terms in lagged y, R, P, CS, CD, NCS, NCD + ε_{i,t}

### Estimation strategy (Arellano-Bond GMM)
- Two-step Arellano-Bond GMM estimator addresses:
  - Endogeneity of global liquidity variables (core and noncore liquidity and their shocks).
  - Predetermined status of inflation and policy rates.
  - Autocorrelation arising from the lagged dependent variable (instrumenting using up to one lag).
  - Fixed effects (time-invariant country characteristics) that may correlate with regressors.

### Additional controls and interaction dummies
- G4 dummy: equals 1 if country is in euro area, Japan, U.K., or U.S., and 0 otherwise (to explore cross-country sensitivity to funding shocks).
- Crisis dummy: equals 1 for observations between Q4 2007 and Q2 2009 (per NBER U.S. recession dating), and 0 otherwise (to test whether the financial crisis altered the impact).
- Post-crisis dummy: equals 1 for period following the financial crisis and 0 otherwise (to assess lingering effects).

---

### Box 4. A VAR Approach to Assessing the Real Impact of Global Liquidity

### Objective and approach
- Use a VAR model similar to the specification in Box 2 to assess real economy impact of liquidity shocks.
- Focus on cumulative impulse response functions of real GDP growth over two, four and eight quarter horizons.
- Cumulative impulse responses approximate level changes in real GDP following a liquidity shock.

### Impulse response computation
- Impulse responses of variable i to a one standard deviation shock in variable j are computed from the structural representation as:
  - IRF_{(i,j,h)} = q' A^h s
  - Where q and s are selection vectors (1 in i-th and j-th elements respectively), Γ is the companion matrix of the VAR, and A_s ... (notation as in the VAR companion form).

---

*Source: Box 1–4, _wp12246 - Estimation of the Noncore Liquidity Price Index and related methodological boxes.*

### Section IV. To establish confidence intervals around the estimates of impulse responses, we

### _wp12246 - Section IV. To establish confidence intervals around the estimates of impulse responses, we

### Bootstrap bias-correction and confidence-interval construction (method)
- Approach: bootstrap confidence intervals with adjustments for small sample sizes as proposed by Kilian (1998).
- Procedure steps:
  - Collect the estimates of the VAR coefficients into a single vector ߚ
መ = vec(ܤ(ܮ)).
  - Randomly resample (with replacement) the residuals from the original estimation and simulate the process with the estimated coefficients ߚ
መ to obtain a thousand of additional samples and coefficient estimates.
  - Denote the average of these additional coefficient estimates by ߚ
כ and estimate the bias of the bootstrap procedure as suggested by Kilian (1998):
    - ߚΔൌ כߚെመ       (5)
  - Construct the bias-corrected estimate ߚ
෨ of the true model parameters ߚ as follows:
    - If all of the eigenvalues of the companion matrix associated with ߚ
መ are inside the unit circle, set ߚ
෨ߚ = መെΔ, otherwise set ߚ
෨ߚ = መ.
    - If any eigenvalues of the companion matrix associated with ߚ
෨ are outside the unit circle, one can further let ߜ
ଵൌ1 and Δ
ଵൌΔ and define Δ
୧ା୨ൌδ
୨Δ and ߜ ௝ାଵ ߜൌ ௝െ0.01. Iteration: set ߚ
෨ߚ = መെΔ
୧ା୨ for ݆ = 1,2,... until all eigenvalues associated with the companion matrix with coefficients ߚ
෨ are inside the unit circle.
  - Second/final step: use ߚ
෨ to generate further additional bootstrap samples, correct them for bias and possible nonstationarity as above, and obtain:
    - the bias-corrected 90 percent confidence interval (using the 5-th and 95-th percentile of the bootstrap distribution of cumulative impulse responses), and
    - the median estimate.
- Outcomes: resulting confidence intervals of the impact of a one standard deviation shock to the supply or demand of core and noncore liquidity are summarized in Figures 12 through 15.

### Notes and methodological caveats
- Alternative approaches mentioned in the source:
  - Monte Carlo integration methods (small sample size is a big limitation).
  - Asymptotically valid confidence intervals using a Bayesian approach (rather than frequentist).
- Stationarity/enforcement device: iterative adjustment when companion-matrix eigenvalues fall outside the unit circle (see above for parameter updates and ߜ/Δ adjustments).

### Key empirical displays and explicit numeric summaries referenced
- Figures referenced for impact estimates and confidence intervals:
  - Figures 12 through 15: impact of Core Demand Shock, Core Supply Shock, Noncore Demand Shock, and Noncore Supply Shock on Real GDP (Change in Level of Real GDP, after 2, 4, and 8 quarters); dashed line denotes subgroup average; 5th and 95th percentiles and median plotted.
- Tables reporting test statistics, trend estimates, and regression results (selected numeric values preserved exactly as in source):
  - Table 1. Unit Root Tests of Liquidity Price and Quantity Indicators, p-values
    - Core liquidity price index: ADF 0.1610, PP 0.4556
    - Noncore liquidity price index: ADF 0.4546, PP 0.6538
    - G4 core liquidity, ratio to GDP: ADF 0.9143, PP 0.8834
    - G4 noncore liquidity, ratio to GDP: ADF 0.5461, PP 0.9991
    - G4 core liquidity, trillion dollars: ADF 0.0465, PP 0.4154
    - G4 noncore liquidity, trillion dollars: ADF 0.8765, PP 0.9305
  - Table 2. Estimation of Linear Trends
    - Variable Core Liquidity, Ratio to GDP: Constant 0.8379***; Time 0.0054***; Adjusted R-squared 0.8201; Durbin Watson statistic 0.1256
    - Variable Noncore Liquidity, Ratio to GDP: Constant 1.0516***; Time 0.0103***; Adjusted R-squared 0.8175; Durbin Watson statistic 0.1210
    - Notes: *** denotes significance at 1 percent.
  - Table 3. Impact of Funding Shocks on Growth: Benchmark Model, Q1 1999–Q1 2011 (selected coefficients and exact statistics as reported)
    - Real GDP growth rate t-1 coefficients: 0.318***, 0.304***, 0.267***, 0.413***, 0.450***, 0.437***, 0.347***, 0.392***, 0.364*** (z-statistics in parentheses corresponding to each estimate are reported in the table).
    - Policy rate t-1 coefficients include -0.212*, -0.134, -0.158, -0.177, -0.118, -0.114, -0.241**, -0.040, -0.197** with z-statistics in parentheses.
    - Headline inflation t coefficients include -0.175***, -0.112, -0.117, -0.158*, -0.136*, -0.131*, -0.069, -0.158***, -0.059 with z-statistics in parentheses.
    - Core liquidity supply shock during crisis t-1: -1.094***, -0.697**, -0.941*, -0.476 (with z-statistics shown).
    - Core liquidity demand shock during crisis t-1: -1.644***, -1.525***, -1.314***, -1.394*** (z-statistics shown).
    - Noncore liquidity supply shock t-1: 0.205***, 0.198, -0.024, 0.214 (z-statistics shown).
    - Noncore liquidity demand shock during crisis t-1: -0.836***, -0.594*, -0.110, -0.024 (z-statistics shown).
    - Constant terms and sample sizes: constants reported (e.g., 5.041***, 4.054***, 4.661***, ...), Number of groups 33 (various), Number of observations 1,192 (as reported).
    - Note: z-statistics in parenthesis. ***,**, * denote significance at 1%, 5% and 10%, respectively. Two-step using Windmeijer standard errors.
  - Table 4. Impact of Funding Shocks on Growth: Separate Time Periods, Q1 1999–Q1 2011
    - Selected coefficients preserved exactly (Real GDP growth rate t-1: 0.305***, 0.302***, 0.296***, 0.420***, 0.452***, 0.435***, 0.341***, 0.392***, 0.328***; Core liquidity supply shock during crisis t-1: -1.022***, -0.545**, -0.587***, -0.677*; etc.).
    - Constant terms and sample sizes as reported (e.g., _cons 5.056***, 4.377***, ...; Number of groups 33; Number of observations 1,192).
    - Note: z-statistics in parenthesis. ***,**, * denote significance at 1%, 5% and 10%, respectively. Two-step using Windmeijer standard errors.
  - Table 5. Impact of Funding Shocks on Growth: G4-Specific Impacts, Q1 1999–Q1 2011
    - Selected coefficients preserved exactly (e.g., Real GDP growth rate t-1: 0.361***, 0.358***, ... across specifications; G4 core supply shock t-1: -0.449***, -0.462***, -0.487***, -0.466***; G4 core demand shock t-1: -0.303***, -0.311***, -0.293***, -0.329***; G4 noncore supply shock t-1: 0.262***, 0.250***, 0.203**, 0.217***; G4 noncore demand shock t-1: -0.543***, -0.546***, -0.561***, -0.566***).
    - Constant terms and sample sizes as reported (e.g., _cons 4.361***, 4.326***, ...; Number of groups 33.000; Number of observations 1,192).
    - Note: z-statistics in parenthesis. ***,**, * denote significance at 1%, 5% and 10%, respectively. Two-step using Windmeijer standard errors.

### Empirical implications presented in this section
- Confidence interval construction:
  - Bias correction via Kilian (1998) bootstrap bias estimate (ߚΔ as above).
  - 90 percent confidence intervals reported for cumulative impulse responses (5-th and 95-th percentiles used).
- Reported impacts (as plotted in Figures 12–15) show country- and subgroup-level median responses and bootstrap 5-th/95-th percentile bands for:
  - Core Demand Shock on Real GDP (changes after 2, 4, 8 quarters).
  - Core Supply Shock on Real GDP (changes after 2, 4, 8 quarters).
  - Noncore Demand Shock on Real GDP (changes after 2, 4, 8 quarters).
  - Noncore Supply Shock on Real GDP (changes after 2, 4, 8 quarters).

*Source: Excerpt from _wp12246 - Section IV (figures and tables referenced within the section).*

### References

### _wp12246 - References

### Liquidity, Global Liquidity, and Credit
- Adrian, Tobias, and Hyun Song Shin, 2010, “Liquidity and Leverage,” Journal of Financial Intermediation, Vol. 19, No. 3, pp. 418–37.  
- Alessi, L. and Detken, C., 2011, “Quasi Real Time Early Warning Indicators for Costly Asset Price Boom/Bust Cycles: A Role for Global Liquidity,” European Journal of Political Economy, Vol 27 (3).  
- Borio, Claudio, R. McCauley, and P. McGuire, 2011, “Global Credit and Credit Booms,” BIS Quarterly Review, September.  
- Borio, Claudio, and P. Disyatat, 2011, “Global Imbalances and The Financial Crisis: Link or No Link?” BIS Working Paper No. 346, May.  
- Borio, Claudio, and Zhu, 2008, “Capital Regulation, Risk-Taking, and Monetary Policy: A Missing Link in the Transmission?” BIS Working Paper No. 268.  
- Bruno, Valentina, and Hyun Song Shin, 2011, “Capital Flows, Cross-Border Banking and Global Liquidity” (forthcoming).  
- Domanski, Dietrich, Ingo Fender, and Patrick McGuire, 2011, “Assessing Global Liquidity,” BIS Quarterly Review, (December).  
- De Nicolo, G., and Ivaschenko, I., 2009, “Global Liquidity, Risk Premiums and Growth Opportunities,” IMF Working Paper 09/52 (Washington: International Monetary Fund).  
- Darius, Reginald, and Sören Radde, 2010, “Can Global Liquidity Forecast Asset Prices?” IMF Working Paper 10/196 (Washington: International Monetary Fund).  
- Committee of Global Financial Stability, 2011, “Global Liquidity—Concept, and Policy Implications,” CGFS Paper 45 (Basel).  
- European Central Bank, 2011, “Global Liquidity: Measurement and Financial Stability Implications,” Financial Stability Review, Special Feature, (December).  
- Rüffer, Rasmus, and Livio Stracca, 2006, “What is Global Excess Liquidity, and Does it Matter?” ECB Working Paper No. 696 (Frankfurt: European Central Bank).  
- McGuire, and von Goetz, 2009, “The US dollar Shortage in Global Banking,” BIS Quarterly Review, pp. 47–63 (March).

### Financial Intermediation, Leverage, and the Financial Sector
- Bernanke, Ben, and Mark Gertler, 1989, “Agency Costs, Net Worth, and Business Fluctuations,” The American Economic Review, Vol. 79, No. 1, pp. 14–31.  
- Bernanke, B, Gertler, M., and Gilchrist, S., “The Financial Accelerator In a Quantitative Business Cycle Framework,” in Handbook of Macroeconomics, Chapter 21, pp. 1341–93, ed. by J. B. Taylor & M. Woodford.  
- Brunnermeier, M., 2009, “Financial Crises: Mechanisms, Prevention, and Management,” in Macroeconomic Stability and Financial Regulation: Key Issues for the G20, ed. by M. Dewartripont, X. Freixas, and R. Portes.  
- Brunnermeier, M., and Yuliy Sannikov, 2011, “A Macroeconomic Model with a Financial Sector” (forthcoming).  
- Gertler, M., and Kiyotaki, N., 2010, “Financial Intermediation and Credit Policy in Business Cycle Analysis,” Handbook of Monetary Economics.  
- Geanakoplos, J., 2010, “The Leverage Cycle,” Cowles Foundation Discussion Paper No. 1715R.  
- Hahm, Joon-Ho, Shin, Hyun Song, and Shin, Kwanho, 2011, “Noncore Bank Liabilities and Financial Vulnerability,” (forthcoming).  
- Poszar, Zoltan, Tobias Adrian, Adam Ashcraft, and Hayley Boesky, 2010, “Shadow Banking,” Federal Reserve Bank of New York, Staff Report No. 458 (July).  
- Poszar, Zoltan, 2011 “Institutional Cash Pools and the Triffin Dilemma of the U.S. Banking System”, IMF Working Paper 11/190 (Washington: International Monetary Fund).  
- Singh, Manmohan, 2011, “Velocity of Pledged Collateral,” IMF Working Paper 11/256 (Washington: International Monetary Fund).  
- Shin, Hyun Song, and Kwanho Shin, 2011, “Pro-cyclicality and Monetary Aggregates,” NBER Working Paper Series No. 16836, February (Cambridge, Massachusetts: National Bureau of Economic Research).  
- Kim, Hyun Jeong, Shin, Hyun Song, and Yun, Jaeho, 2012, “Monetary Aggregates and the Central Bank’s Financial Stability Mandata”, forthcoming

### IMF Analyses, Reports, and Technical Notes
- International Monetary Fund, 2011a, “How to Address the Systemic Part of Liquidity Risk,” Global Financial Stability Report, Chapter 2 (April), World Economic and Financial Surveys (Washington).  
- _________, 2011b, “Global Liquidity Management—Possible Indicators to Monitor Global Liquidity,” Background Paper (SM/11/277) (Washington).  
- _________, 2011c, “The United States Spillover Report—2011 Article IV Consultation,” IMF Country Report No. 11/203. Available via Internet: http://www.imf.org/external/pubs/cat/longres.aspx?sk=25083.0  
- _________, 2011d, “The United Kingdom Spillover Report—2011 Article IV Consultation,” IMF Country Report No. 11/225. Available via Internet: http://www.imf.org/external/pubs/cat/longres.aspx?sk=25114.0  
- _________, 2011e, “How to Address the Systemic Part of Liquidity Risk,” Global Financial Stability Report, Chapter 2 (April), World Economic and Financial Surveys (Washington: International Monetary Fund).  
- _________, 2011f, “Technical Note For G20 Sub-Working Group on Measuring Global Liquidity,” March, mimeo.  
- _________, 2011g, “Analytics of Systemic Crises and the Role of Global Financial Safety Nets,” SM/11/107.  
- _________, 2011h, “Mapping Cross-Border Financial Linkages—A Supporting Case for Global Financial Safety Nets,” SM/11/108.  
- _________, 2011i, “Housing Finance and Financial Stability—Back to Basics?,” Global Financial Stability Report, Chapter 3 (April), World Economic and Financial Surveys (Washington: International Monetary Fund).  
- _________, 2011j, “Technical Note For G20 Sub-Working Group on Global Liquidity: Progress Report on Assessing Global Liquidity,” June, mimeo.  
- _________, 2010a, “Global Liquidity Expansion: Effects on ‘Receiving’ Economies and Policy Response Options,” in Global Financial Stability Report, Chapter 4 (April), World Economic and Financial Surveys (Washington: International Monetary Fund).  
- _________, 2010b, “Systemic Liquidity Risk: Improving the Resilience of Financial Institutions and Markets,” in Global Financial Stability Report, Chapter 2 (October), World Economic and Financial Surveys (Washington: International Monetary Fund).  
- _________, 2009, “Responding to the Financial Crisis and Measuring Systemic Risks,” in Global Financial Stability, Chapter 1 (April), World Economic and Financial Surveys (Washington: International Monetary Fund).  
- _________, 2009, “Lessons for Monetary Policy from Asset Price Fluctuations,” in World Economic Outlook, Chapter 3 (October), World Economic and Financial Surveys (Washington: International Monetary Fund).  
- _________, 2008, “Containing Systemic Risks and Restoring Financial Soundness,” in Global Financial Stability, Chapter 3 (April), World Economic and Financial Surveys (Washington: International Monetary Fund).  
- _________ , 2007, “What is Global Liquidity?” Box 1.3 in World Economic Outlook, Chapter 1 (October), World Economic and Financial Surveys (Washington: International Monetary Fund).

### Monetary Policy, Business Cycles, and Asset Prices
- Canova, F., and De Nicolo, G., 2002, “Monetary Disturbances Matter for Business Fluctuations in the G-7,” Journal of Monetary Economics, Vol. 49, pp. 1131–59.  
- D’Agostino, A., and Surico, P., 2009, “Does Global Liquidity Help to Forecast U.S. Inflation?” Journal of Money, Credit, and Banking, Vol. 41, pp. 479–89.  
- Fry, R., and Pagan, A., 2011, “Sign Restrictions in Structural Vector Autoregressions: A Critical Review,” Journal of Economic Literature (forthcoming).  
- Giannone, D., Reichlin L., and Small, D., 2008, “Nowcasting: The Real-time Informational Content of Macroeconomic Data,” Journal of Monetary Economics, Vol. 55(4), pp. 665–76.  
- Gerdesmeier, D., Reimers, H.-E. and Roffia, B., 2010, “Asset Price Misalignments and the Role of Money and Credit”, International Finance, Vol. 13, pp. 377–407.  
- Matheson, T., 2011, “Financial Conditions Indexes for the United States and Euro Area,” IMF Working Paper 11/93 (Washington: International Monetary Fund).  
- Peersman, G., 2005 “What Caused the Early Millennium Slowdown? Evidence based on Vector Autoregressions,” Journal of Applied Econometrics, Vol. 20 (2), pp. 185–207.  
- Sousa, João, and Andrea Zaghini, 2004, “Monetary Policy Shocks in the Euro Area and Global Liquidity Spillovers,” Bank of Italy Economic Working Papers No. 629.  
- Thomas, Alun, Martin Mühleisen, and Malika Pant, 2010, “Peaks, Spikes and Barrels: Modeling Sharp Movements in Oil Prices,” IMF Working Paper 10/186 (Washington: International Monetary Fund).  
- Uhlig, H., 2005, ‘What are the Effects of Monetary Policy on Output? Results from an Agnostic Identification Procedure,” Journal of Monetary Economics, Vol. 52, pp. 381–419.

### Econometric and Theoretical Methods
- Kilian, Lutz, 1998, “Small-Sample Confidence for Impulse Response Functions,” The Review of Economics and Statistics, Vol. 80 (2), pp. 218–30.  
- Kilian, L., 2009, “Not All Oil Price Shocks Are Alike: Disentangling Demand and Supply Shocks in the Crude Oil Market,” American Economic Review, Vol. 99:3, pp. 1053–69.  
- Schwarz, Gideon, 1978, “Estimating the Dimension of a Model,” The Annals of Statistics 6(2), pp. 461–64.  
- Tirole, Jean, 2011, “Illiquidity and All its Friends,” Journal of Economic Literature, Vol. 49, Number 2, pp. 287–325.  
- Jorda, Oscar, Schularick Moritz, Alan M. Taylor, 2011, “When Credit Bites Back: Leverage, Business Cycles, and Crises,” Federal Reserve Bank of San Francisco, Working Paper Series, (November).  
- Kiyotaki, N., and Moore, J., 1997, “Credit Cycles,” The Journal of Political Economy, Vol. 105, No. 2, pp. 211–48.  
- Eickmeier, S., and Ng, T., 2011, “How do Credit Supply Shocks Propagate Internationally? A GVAR Approach,” CEPR Discussion Paper No. 8720, (December).  
- Kilian, Lutz, 1998, “Small-Sample Confidence for Impulse Response Functions,” The Review of Economics and Statistics, Vol. 80 (2), pp. 218–30.

*References list from _wp12246 - References (source PDF).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp12246.pdf_