## _wp0952

## Source details

**Canonical URL:** [_wp0952](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp0952.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp0952.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp0952.pdf.json)

---

### Introduction
- Paper constructs market liquidity indicators and assesses their role as priced risk factors and determinants of price-earnings (PE) ratios.
- Threefold contribution:
  - Construct new market liquidity indicators for equity, bond, and money markets using price data and time-varying volatility models; build global systemic liquidity shock measures.
  - Assess liquidity indicators as risk factors embedded in bond spreads for advanced economies (bond yield spreads over the risk-free rate) and EMBI spreads for emerging markets.
  - Assess whether liquidity indicators are positively associated with proxy measures of countries’ growth opportunities (price-earnings ratios).
- Three main empirical results:
  - Market liquidity increased in virtually all markets and countries since at least the mid-1990s; strongest in emerging markets; cross-country correlations of liquidity indicators increased; incidence of global systemic liquidity shocks increased.
  - Liquidity indicators (national and U.S.) are important determinants of bond premiums (bond yield spreads and EMBI spreads).
  - National and U.S. equity market liquidity indicators are significant determinants of price-earnings ratios in many countries.

### Liquidity indicators: theory and measurement
- Theoretical framework (microstructure-based; Hasbrouck (2007) extension, Roll (1984), Kyle (1985)):
  - Fundamental price law of motion: 1tt tt tt VVqλσε − =++ (equations preserved conceptually in source).
  - Trading price: tttt PVcq=+.
  - Trade direction q_t ∈ {+1 (sell/ask), −1 (buy/bid)} assumed equal probability.
  - λ_t captures illiquidity / adverse selection; c_t captures non-informational transaction costs.
  - Illiquidity generates first-order autocovariance patterns (Roll (1984) case: negative first-order autocorrelation); momentum/contrarian trading induce non-zero higher-order autocovariances.
- Measurement desiderata:
  1. Rooted in finance theory.
  2. Easily computable across asset classes using readily available data.
  3. Capture market-wide liquidity.
- Empirical construction (monthly horizon):
  - Daily returns modeled with GARCH(1,1) to allow time-varying volatility; monthly variance includes variances and covariances of daily returns.
  - Liquidity indicator L_t (monthly) defined as ratio:
    - Numerator: sum of absolute values of negative covariances and positive covariances of daily returns within the month (2 * sum of |covariances| as structured in source).
    - Denominator: numerator plus sum of variances of daily returns within the month.
    - Range: unit interval; the smaller L_t is, the higher is liquidity.
  - Novel aspects: captures entire autocovariance structure, allows time-varying variances/covariances via GARCH, simple price-data-only computation suitable for cross-country comparisons.

### Liquidity indicators: evidence
- Data coverage:
  - Value-weighted price indices for a sample of 30 countries: G-7, five Australasian industrial countries, a group of emerging markets, and global aggregate.
  - Daily and monthly data from January 1, 1980 to April 31, 2008 on broad stock indices, government bond indices, and money market indices for industrial countries; bond yields and spreads from proprietary sources.
  - Typical series start dates: stock markets January 1980; industrial bond markets January 1985; emerging bond markets and spreads between January 1988 and March 1993.
- Descriptive patterns:
  - Global liquidity indicators for equity, bond, and money markets show significant fluctuations, align with known turbulence episodes, and capture recent contraction of market liquidity.
  - Advanced economies generally exhibit greater liquidity and lower volatility of liquidity than emerging markets.
  - Within advanced economies: stock market liquidity < government bond market liquidity < money market liquidity.
- Dynamics and co-movements:
  - Log liquidity autoregressive-with-trend model Ln L_t = α + β t + γ Ln L_{t−1} + ε used.
  - Negative time-trend coefficients indicate liquidity increased in almost all markets since mid-1990s; most pronounced in emerging markets (exceptions: Russia and Turkey).
  - Liquidity indicators positively correlated across domestic markets and with global liquidity indicators; linkages strengthened over time.
  - Cross-country variance declines (especially for stock markets) imply increased correlations; declines in cross-country variance of government bond liquidity predict declines in cross-country variance of equity liquidity.
  - Interpretation: increased financial integration contributed to stronger global and cross-market liquidity co-movements.
- Systemic liquidity shocks:
  - Systemic shock indicator: percentage of countries with national liquidity indicator in the upper 10th percentile of their historical distribution (constructed separately for stock and bond markets).
  - Incidence of global liquidity shocks increased recently across stock and bond markets.
  - Contemporaneous correlation between stock- and bond-market systemic liquidity shock measures: 0.55.
  - Bi-variate VAR(1): a systemic liquidity shock to equity markets leads to a decline of liquidity in bond markets, but not vice versa — transmission from equity-market liquidity shocks to bond-market liquidity.

### Liquidity and bond premiums
- Concept: liquidity measures capturing trading frictions related to asymmetric information and transaction costs should be priced and affect spreads; global (U.S.) liquidity components may drive spreads internationally.
- Advanced economies (Government bond spreads; dependent variable: first difference of spread between a long term yield and a short term interest rate; fixed effects panel and OLS specifications):
  - Key coefficient L.spread across columns (1)–(11):
    - (1) -0.0801*** [0.00275]
    - (2) -0.0773*** [0.00875]
    - (3) -0.0665*** [0.00184]
    - (4) -0.145*** [0.000198]
    - (5) -0.0488** [0.0252]
    - (6) -0.216*** [0.00569]
    - (7) -0.0770* [0.0652]
    - (8) -0.203*** [0.000100]
    - (9) -0.0639* [0.0639]
    - (10) -0.0756 [0.118]
    - (11) -0.0271 [0.321]
  - US Bond Liquidity (selected columns):
    - (1) 4.301*** [0.000570]
    - (2) 4.413** [0.0301]
    - (3) 4.689*** [0.0000145]
    - (4) 13.77*** [0.00456]
    - (5) 5.558** [0.0188]
    - (6) 7.705** [0.0439]
    - (7) 0.930* [0.0949]
    - (8) 2.008 [0.219]
    - (9) 5.379*** [0.00149]
  - US Stock Liquidity (selected columns):
    - (1) -231.1** [0.0327]
    - (2) -689.6 [0.139]
    - (3) -204.3 [0.329]
  - US Money Market Liquidity (selected columns):
    - (1) -0.187* [0.0912]
    - (7) -0.240*** [0.00409]
    - (9) -0.639*** [0.00507]
  - Summary of Table 4 findings:
    - Panel regression with domestic liquidity indicators only: none significant.
    - Adding U.S. liquidity indicators: U.S. bond market liquidity is significantly associated with declines in bond spreads across countries.
    - U.S. equity and money market liquidity impacts on bond spreads are negative and significant in the panel — interpreted as portfolio substitution effects.
    - Country-by-country regressions: declines in bond spreads associated with improvements in U.S. bond market liquidity in most countries; domestic money market liquidity significantly impacts bond spreads in most countries.
  - Sample and fit:
    - Observations by column: 1483, 1479, 208, 208, 208, 99, 173, 205, 149, 115, 114
    - R-squared by column: 0.045, 0.132, 0.12, 0.222, 0.273, 0.326, 0.157, 0.161, 0.088, 0.276, 0.367
- Emerging economies (EMBI spreads; dependent variable: first difference of natural logarithm of EMBI spread; fixed effects panel with clustered standard errors):
  - Key coefficients across columns (1) All, (2) Latin America, (3) Asia, (4) Europe:
    - L.dlembi:
      - (1) 0.126* [0.0847]
      - (2) 0.132 [0.233]
      - (3) 0.109** [0.0118]
      - (4) 0.107 [0.214]
    - Stock Liquidity:
      - (1) 0.0357* [0.0567]
      - (2) 0.0539* [0.0936]
      - (3) 0.0620* [0.0524]
      - (4) -0.0345 [0.344]
    - US Bond Liquidity:
      - (1) -1.044** [0.0249]
      - (2) -1.781 [0.135]
      - (3) -0.564** [0.0180]
      - (4) -0.552 [0.219]
    - US Stock Liquidity:
      - (1) 0.176*** [0.000800]
      - (2) 0.243** [0.0151]
      - (3) 0.133** [0.0233]
      - (4) 0.139 [0.154]
  - Sample and fit:
    - Observations: 1028, 394, 416, 218
    - # of countries: 16, 6, 6, 4
    - R-squared: 0.055, 0.078, 0.047, 0.03
  - Summary of Table 5 findings:
    - Domestic equity liquidity SL: positive and significant association with lower EMBI spreads.
    - Improvements in U.S. equity market liquidity significantly associated with lower EMBI spreads.
    - Worsening of U.S. bond market liquidity relative to domestic liquidity associated with a decline in EMBI spreads (negative and significant US bond liquidity coefficient).

### Liquidity and growth opportunities (PE ratios)
- Hypothesis: more liquid equity markets lower firms’ cost of capital and enhance growth opportunities; PE ratios used as proxy for domestic growth opportunities.
- Model: first differences in PE ratios regressed on domestic equity liquidity SL, U.S. equity liquidity SLUS, world PE ratio PEW, lags and country fixed effects.
- Table 6 key coefficients across columns (1) All, (2) G-7, (3) All Developed, (4) Emerging, (5) Emerging Latin America, (6) Emerging Asia, (7) Emerging Europe:
  - L.pe:
    - (1) -0.296*** [0.000]
    - (2) -0.0472*** [0.001]
    - (3) -0.0537*** [0.000]
    - (4) -0.334*** [0.000]
    - (5) -0.363*** [0.000]
    - (6) -0.128*** [0.000]
    - (7) -0.109** [0.024]
  - World PE ratio:
    - (1) 0.168*** [0.004]
    - (2) 0.0724** [0.014]
    - (3) 0.0692*** [0.000]
    - (4) -0.0493 [0.452]
    - (5) -0.156 [0.238]
    - (6) 0.0553* [0.065]
    - (7) 0.0259 [0.564]
  - Stock Market Liquidity (domestic):
    - (1) 1.242 [0.518]
    - (2) -0.866* [0.055]
    - (3) -0.797*** [0.004]
    - (4) 3.946 [0.245]
  - US Stock Market Liquidity:
    - (1) -3.624** [0.025]
    - (2) -1.411*** [0.003]
    - (3) -1.421*** [0.000]
    - (4) -5.204** [0.018]
    - (5) -7.579* [0.081]
    - (6) -2.831** [0.013]
    - (7) -0.856* [0.098]
  - Sample and fit:
    - Observations: 6931, 2241, 3715, 3216, 1060, 1384, 772
    - # of countries: 30, 7, 12, 18, 6, 7, 5
    - R-squared: 0.175, 0.064, 0.062, 0.196, 0.212, 0.077, 0.058
- Interpretation:
  - Full sample: U.S. equity market liquidity has a significant positive impact on PE ratios (improvement in liquidity → increase in PE; note liquidity indicator scaling in model implies negative coefficient on liquidity indicator corresponds to positive liquidity effect).
  - Advanced economies: both domestic and U.S. equity market liquidity have significant positive impacts on PE ratios.
  - Emerging economies: U.S. equity market liquidity retains a positive impact on PE ratios; domestic equity liquidity is not significant in emerging samples.
  - Conclusion: improvements in global equity market liquidity foster national growth opportunities; domestic liquidity improvements contribute positively primarily in advanced economies.

### Conclusions and extensions
- Measures and methodology:
  - Novel liquidity measures for equity, bond, and money markets based on mapping liquidity into autocovariance structure of returns; applied to major advanced and emerging market countries.
- Historical findings:
  - Documented historical increase in market liquidity, likely due to advances in international financial integration.
  - Financial markets increasingly exposed to global systemic liquidity shocks.
  - Liquidity indicators are important determinants of bond spreads and proxy measures of countries’ growth opportunities.
- Empirical relationships:
  - Liquidity measures linked to bond spreads and PE ratios; cross-sectional dynamics show interactions between equity and bond market liquidity variances.
- Extensions and open questions:
  - Compare these measures with other liquidity measures; extend to other markets and assets within markets.
  - Open question: whether recent turmoil represents a temporary reversal or a more permanent break in the progress in market liquidity and international financial integration.

*Source: _wp0952 (IMF working paper chapter and tables provided in the content unit).*

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

### _wp0952 - References..............................................................................................................

### Figures
- 1.  Global Liquidity Indicators ................................................................................................19
- 2. G-7 Equity and Bond Liquidity Indicators ..........................................................................20
- 3. Emerging Markets Liquidity Indicators: Latin America .....................................................21
- 4. Emerging Markets Liquidity Indicators: Asia .....................................................................22
- 5. Emerging Markets Liquidity Indicators: Europe .................................................................23
- 6. Liquidity Indicators: Mean and Standard Deviation............................................................24
- 7. Liquidity Indicators: Trend Coefficients .............................................................................26
- 8. Liquidity Indicators: Correlations........................................................................................27
- 9. Indicators of Global Systemic Liquidity Shocks .................................................................29

### Tables
- 1. Fixed Effect Panel Regressions ...........................................................................................25
- 2. Liquidity Indicators: Dynamics of Cross Sectional Variances ............................................28
- 3. Bi-variate VAR of Global Equity and Bond Systemic Liquidity Shocks............................30
- 4. Government Bond Spreads and Liquidity: Advanced Economies ......................................31
- 5. EMBI Spreads and Liquidity: Emerging Economies...........................................................32

*Source: _wp0952 - References..............................................................................................................*

### 6. Price-Earning (PE) Ratios and Liquidity .............................................................................

### 6. Price-Earning (PE) Ratios and Liquidity

### Introduction
- Popular notion: the “global economy is awash with liquidity” interpreted as abundance of money and “excess liquidity”, conjectured to be driven by accommodative monetary policies and to have contributed to the decline of risk premiums; mid-August 2007 saw a sudden apparent reversal described as a “liquidity” or “credit” crunch.
- Definitional and modeling challenges:
  - Multiple, often vague notions of liquidity in press and policy discourse.
  - Integration of liquidity into general equilibrium macro models is at an early stage.
  - Finance literature has progressed faster in modeling and measuring liquidity premiums and liquidity risk, but most empirical work focuses on U.S. markets.
- Paper’s contribution (threefold):
  1. Construct new market liquidity indicators rooted in finance theory that use only readily available market price data and capture dynamics of liquidity shocks via returns models with time-varying volatility; indicators cover equity, bond, and money markets for major advanced and emerging market countries, at national and global levels; construct measures of global systemic liquidity shocks and document their evolution and joint dynamics.
  2. Assess to what extent liquidity indicators (national and global) are risk factors embedded in bond spreads for advanced economies (bond yield spreads over the risk-free rate) and emerging markets (EMBI spreads).
  3. Assess whether liquidity indicators are positively associated with proxy measures of countries’ growth opportunities (price-earnings ratios).
- Three main empirical results:
  - Market liquidity increased in virtually all markets and countries since at least the mid-1990s, with most pronounced advances in emerging markets; cross-country correlations of liquidity indicators increased; incidence of global systemic liquidity shocks increased.
  - Liquidity indicators (national and U.S.) are important determinants of bond premiums (bond yield spreads and EMBI spreads).
  - National and U.S. equity market liquidity indicators are significant determinants of price-earnings ratios in many countries; liquidity is a channel through which financial integration can translate into higher real growth.

*Italicized source attribution line.*

### Liquidity indicators: theory and measurement
- Theoretical framework:
  - Microstructure-based model (Hasbrouck (2007) extension, Roll (1984), Kyle (1985)) with fundamentals Vt (log) and execution price Pt (log):
    - Fundamental price law of motion: 1tt tt tt VVqλσε − =++  (equation numbers preserved from source).
    - Trading price: tttt PVcq=+.
    - Trade direction q_t ∈ {+1 (sell/ask), −1 (buy/bid)} assumed equal probability.
    - λ_t captures illiquidity / adverse selection (Kyle (1985)); c_t captures non-informational transaction costs.
  - Returns and variance/covariance structure:
    - Daily/monthly return expressions and autocovariances show how liquidity (via λ_t and c_t) maps into return autocovariances and variances (equations (3), (4), (5) preserved conceptually).
  - Key implications:
    - Illiquidity generates first-order autocovariance patterns (Roll (1984) case: negative first-order autocorrelation).
    - Momentum or contrarian trading (non-zero lagged trade direction correlations) induce non-zero higher-order autocovariances; evolution of autocovariances tracks liquidity and information/revelation dynamics.

- Measurement desiderata:
  1. Rooted in finance theory.
  2. Easily computable across many asset classes using readily available data.
  3. Capture market-wide liquidity (not idiosyncratic asset liquidity).

- Empirical construction (monthly horizon):
  - Investment horizon divided into K trading days; monthly return equals sum of daily log returns.
  - Daily returns modeled with GARCH(1,1) to allow time-varying volatility; monthly variance includes variances and covariances of daily returns (equation (6) preserved conceptually).
  - Liquidity indicator L_t (monthly) defined as ratio:
    - Numerator: sum of absolute values of negative covariances and positive covariances of daily returns within the month (2 * sum of |covariances| as structured in source).
    - Denominator: numerator plus sum of variances of daily returns within the month (expression in source preserved conceptually as equation (7)).
    - Range: unit interval; the smaller L_t is, the higher is liquidity.
  - Novel aspects:
    - Captures entire autocovariance structure of within-period returns (including positive and negative covariances).
    - Allows time-varying variances and covariances via GARCH.
    - Simple, price-data-only computation applicable to price indexes and suitable for cross-country, cross-market comparisons—advantageous in emerging markets where tick/cluster data are limited.

### Liquidity indicators: evidence
- Data:
  - Value-weighted price indices for a sample of 30 countries: G-7, five Australasian industrial countries, a group of emerging markets, and global aggregate.
  - Daily and monthly data from January 1, 1980 to April 31, 2008 on broad stock indices, government bond indices, and money market indices for industrial countries; bond yields and spreads from proprietary sources.
  - Primary sources: DataStream (JP Morgan bond indices and Citigroup cash indices) for stock markets, government bond and money markets for industrial countries; Bloomberg (JP Morgan EMBI and constituents) for emerging bond market returns, yields, and spreads.
  - Typical series start dates: stock markets January 1980; industrial bond markets January 1985; emerging bond markets and spreads between January 1988 and March 1993.
  - Monthly inflation and interest rates: IFS; P/E ratios: DataStream.

- Descriptive statistics and ranking:
  - Global liquidity indicators for equity, bond, and money markets show significant fluctuations, align with known market turbulence episodes, and capture the recent contraction of market liquidity.
  - Cross-country patterns:
    - Advanced economies generally exhibit greater liquidity and lower volatility of liquidity than emerging markets.
    - Within markets: stock market liquidity < government bond market liquidity < money market liquidity in advanced economies.
  - Figures referenced show timing precision and cross-market patterns (Figures 1–6 in source).

- Dynamics and co-movements:
  - Time-series regression for log liquidity as autoregressive with trend: Ln L_t = α + β t + γ Ln L_{t−1} + ε (equation (8) concept).
  - Findings:
    - Negative time-trend coefficients in panel regressions indicate liquidity increased in almost all markets since mid-1990s, most pronounced in emerging markets (exceptions: Russia and Turkey).
    - Liquidity indicators positively correlated across domestic markets and with global liquidity indicators; linkages strengthened over time.
  - Cross-sectional variance model for co-movements (equation (9) preserved conceptually): results indicate:
    - Declining cross-country variance of liquidity (particularly for stock markets) implying increased correlations.
    - Evidence that declines in cross-country variance of government bond liquidity predict declines in cross-country variance of equity market liquidity — suggesting mutually reinforcing co-movements across connected markets.
  - Interpretation: increased financial integration likely contributed to stronger global and cross-market liquidity co-movements.

- Systemic liquidity shocks:
  - Systemic shock indicator definition: percentage of countries with national liquidity indicator in the upper 10th percentile of their historical distribution (constructed separately for stock and bond markets).
  - Empirical findings:
    - Dynamics show incidence of global liquidity shocks increased recently across stock and bond markets (Figure 9 referenced).
    - Contemporaneous correlation between stock- and bond-market systemic liquidity shock measures: 0.55.
    - Bi-variate VAR(1) estimates (Table 3): a systemic liquidity shock to equity markets leads to a decline of liquidity in bond markets, but not vice versa — indicating transmission from equity-market liquidity shocks to bond-market liquidity (spillover from flight-to-quality to flight-to-liquidity interpretation).

### Liquidity and bond premiums
- Conceptual premise:
  - If liquidity measures capture trading frictions related to asymmetric information and transaction costs, they should represent priced risk factors affecting spreads.
  - Global (U.S.) liquidity components may drive spreads internationally due to market integration and common liquidity components.

- Advanced economies (long-term government bond yield minus three-month rate = BSPREAD):
  - Estimated fixed-effect dynamic panel regression (equation (10) in source) with domestic liquidity indicators BL (bond), SL (equity), CL (money), U.S. liquidity indicators (BLUS, SLUS, CLUS), US bond spread USBSPREAD, lags, and country fixed effects.
  - Main empirical findings (Table 4 referenced):
    - Panel regression with domestic liquidity indicators only: none significant.
    - Adding U.S. liquidity indicators: U.S. bond market liquidity (decline in liquidity indicator = improvement) is significantly associated with declines in bond spreads across countries.
    - U.S. equity and money market liquidity impacts on bond spreads are negative and significant in the panel — an improvement in U.S. equity or money market liquidity, holding other markets constant, is associated with an increase in another country’s bond spread (interpreted as portfolio substitution effects).
    - Country-by-country regressions confirm panel results: declines in bond spreads associated with improvements in U.S. bond market liquidity in most countries (Japan exception: coefficient positive but not significant); domestic money market liquidity significantly impacts bond spreads in most countries.
  - Summary: U.S. bond liquidity and domestic money market liquidity are significant risk factors embedded in bond spreads — evidence of increased liquidity interconnectedness.

- Emerging economies (EMBI spreads):
  - Model: first differences of log EMBI spreads (Δ ln EMBI) regressed on domestic equity liquidity SL, U.S. liquidity measures (BLUS, SLUS, CLUS) and country fixed effects (equation (11) in source).
  - Findings (Table 5 referenced):
    - Positive and significant coefficient on domestic equity liquidity SL: increased domestic liquidity associated with lower EMBI spreads.
    - Improvements in U.S. equity market liquidity significantly associated with lower EMBI spreads (positive and significant coefficient for US equity market liquidity measure in source formulation).
    - Worsening of U.S. bond market liquidity relative to domestic liquidity is significantly associated with a decline in EMBI spreads (negative and significant coefficient on U.S. bond liquidity indicator).
  - Summary: both domestic and global (U.S.) liquidity indicators are important risk factors embedded in EMBI spreads; role of U.S. liquidity consistent with increased sensitivity of risk premiums to global factors.

### Liquidity and growth opportunities (PE ratios)
- Hypothesis: more liquid equity markets lower firms’ cost of capital and enhance growth opportunities; P/E ratios proxy domestic growth opportunities.
- Empirical model: first differences in PE ratios regressed on domestic equity liquidity SL, U.S. equity liquidity SLUS, world PE ratio PEW, lags and country fixed effects (equation (12) in source).
- Main findings (Table 6 referenced):
  - Full sample panel: U.S. equity market liquidity has a significant positive impact on PE ratios (improvement = decline in liquidity indicator → increase in PE); domestic liquidity not significant.
  - Advanced economies (G-7 and all advanced): both domestic and U.S. equity market liquidity have significant positive impacts on PE ratios.
  - Emerging economies: U.S. equity market liquidity retains a positive impact on PE ratios; domestic equity liquidity is not significant in emerging samples.
- Interpretation:
  - Improvements in global equity market liquidity foster national growth opportunities.
  - Domestic equity liquidity improvements contribute positively to growth opportunities primarily in advanced economies.
  - Findings complement earlier results that financial integration fosters equity market liquidity and suggest liquidity is a channel through which financial integration lowers cost of capital and raises real growth opportunities.

### Conclusions (as indicated up to Section VI)
- Empirical evidence in the paper supports three core conclusions:
  - Market liquidity increased in most countries and markets since at least the mid-1990s, with stronger cross-country co-movements and increased incidence of systemic liquidity shocks; emerging markets experienced particularly pronounced liquidity improvements.
  - Both national and U.S. liquidity indicators are important determinants (priced risk factors) of bond spreads in advanced economies and EMBI spreads in emerging markets; U.S. liquidity measures play a key global role.
  - Equity market liquidity (especially global/U.S. liquidity and, in advanced economies, domestic liquidity) is positively associated with price-earnings ratios, implying liquidity as a channel linking financial integration to reduced cost of capital and higher growth opportunities.

*Italicized source attribution line.*

### CONCLUSION

### CONCLUSION

### Measures and Methodology
- Constructed novel measures of liquidity for equity, bond and money markets in major advanced and emerging market countries based on a simple model that maps the evolution of market liquidity into the autocovariance structure of asset returns.
- Liquidity indicators derived from the model are applied to equity, government bond, and money markets.

### Historical Findings
- Documented a historical increase in market liquidity likely due to advances in international financial integration.
- At the same time, financial markets have been increasingly exposed to global systemic liquidity shocks.
- The liquidity indicators appear to be important determinants of bond spreads and of proxy measures of countries’ growth opportunities.

### Empirical Relationships and Dynamics
- Liquidity measures are linked empirically to:
  - Bond spreads.
  - Proxy measures of countries’ growth opportunities.
- Cross-sectional dynamics show interactions between equity and bond market liquidity variances, with models estimating relationships among lagged variances and trends (see reported panel estimates and VAR analyses in the chapter for detailed coefficients and statistical significance).

### Extensions, Applications, and Open Questions
- The measures could be compared with other liquidity measures.
- Possible extensions include:
  - Extending the measures to other markets.
  - Applying the measures to different assets within the same markets.
- These extensions could provide useful insights about the evolution of market liquidity as an important facet of financial development.
- A pressing question for future research: whether the recent turmoil in financial markets represents just a temporary reversal of the progress in market liquidity witnessed in the past decade, or a more permanent break in international financial development and integration.

*Source: CONCLUSION, _wp0952 - CONCLUSION*

### 1. Constant                  0.017615241  0.006004648      2.93360  0.00335055 2. SLSSTOCK{1}               0.817907057 

### Tables 4–6: Liquidity and Asset Spreads / PE Ratios

### Government Bond Spreads and Liquidity: Advanced Economies (Table 4)
- Dependent variable: the first difference of the spread between a long term yield and a short term interest rate on government bonds. Estimation: fixed effects panel regressions for (1)-(2) (standard errors clustered by country) and OLS for (3)-(11). Robust p-values in brackets. *, **, *** denote significance at 10%, 5% and 1% levels respectively.
- COEFFICIENT L.spread (columns (1) to (11)):
  - (1) -0.0801*** [0.00275]
  - (2) -0.0773*** [0.00875]
  - (3) -0.0665*** [0.00184]
  - (4) -0.145*** [0.000198]
  - (5) -0.0488** [0.0252]
  - (6) -0.216*** [0.00569]
  - (7) -0.0770* [0.0652]
  - (8) -0.203*** [0.000100]
  - (9) -0.0639* [0.0639]
  - (10) -0.0756 [0.118]
  - (11) -0.0271 [0.321]
- Bond Liquidity (columns (1) to (11)):
  - (1) -0.157 [0.616]
  - (2) -0.269 [0.330]
  - (3) 6.539*** [0.000336]
  - (4) -0.21 [0.741]
  - (5) -0.0772 [0.770]
  - (6) -2.620* [0.0528]
  - (7) 0.0797 [0.922]
  - (8) 1.202 [0.203]
  - (9) 0.164 [0.333]
  - (10) 0.003 [0.993]
  - (11) -0.169 [0.640]
- Stock Liquidity (columns (1) to (11)):
  - (1) -0.205 [0.149]
  - (2) -0.226 [0.203]
  - (3) -0.364* [0.0511]
  - (4) -1.359*** [0.00264]
  - (5) -0.267 [0.162]
  - (6) -0.197 [0.823]
  - (7) -0.149 [0.650]
  - (8) -0.444 [0.428]
  - (9) 0.000325 [0.997]
  - (10) -0.139 [0.696]
  - (11) 0.808** [0.0296]
- Money Market Liquidity (columns (1) to (11)):
  - (1) -0.0165 [0.926]
  - (2) -0.0714 [0.558]
  - (3) 103.8 [0.678]
  - (4) -89.29 [0.386]
  - (5) -279.8** [0.0231]
  - (6) -113.8** [0.0397]
  - (7) 7.649** [0.0465]
  - (8) -563.1** [0.0228]
  - (9) 0.00352 [0.973]
  - (10) -1747** [0.0117]
  - (11) 135.5 [0.731]
- US Bond Liquidity (selected columns shown):
  - (1) 4.301*** [0.000570]
  - (2) 4.413** [0.0301]
  - (3) 4.689*** [0.0000145]
  - (4) 13.77*** [0.00456]
  - (5) 5.558** [0.0188]
  - (6) 7.705** [0.0439]
  - (7) 0.930* [0.0949]
  - (8) 2.008 [0.219]
  - (9) 5.379*** [0.00149]
- US Stock Liquidity (selected columns shown):
  - (1) -231.1** [0.0327]
  - (2) -689.6 [0.139]
  - (3) -204.3 [0.329]
  - (4) -73.06 [0.935]
  - (5) -335.1 [0.377]
  - (6) 339.3 [0.520]
  - (7) -123.1 [0.338]
  - (8) -213.6 [0.542]
  - (9) 22.98 [0.946]
- US Money Market Liquidity (selected columns shown):
  - (1) -0.187* [0.0912]
  - (2) 0.257 [0.468]
  - (3) -0.273 [0.139]
  - (4) -0.741 [0.475]
  - (5) -0.171 [0.577]
  - (6) -0.772 [0.117]
  - (7) -0.240*** [0.00409]
  - (8) -0.168 [0.527]
  - (9) -0.639*** [0.00507]
- US Spread (selected columns shown):
  - (1) 0.256** [0.0205]
  - (2) 0.272** [0.0272]
  - (3) 0.120** [0.0120]
  - (4) 0.231 [0.234]
  - (5) -0.0635 [0.490]
  - (6) 0.333** [0.0211]
  - (7) 0.0201 [0.492]
  - (8) 0.230*** [0.00244]
  - (9) 0.296*** [0.0000849]
- Constant (columns (1) to (11)):
  - (1) 0.133* [0.0584]
  - (2) 0.101* [0.0971]
  - (3) -0.0516 [0.328]
  - (4) 0.487*** [0.000362]
  - (5) 0.133** [0.0102]
  - (6) 0.512** [0.0400]
  - (7) -0.00819 [0.931]
  - (8) 0.0602 [0.660]
  - (9) 0.0865 [0.187]
  - (10) 0.291*** [0.00571]
  - (11) -0.132 [0.308]
- Sample and fit:
  - Observations: 1483, 1479, 208, 208, 208, 99, 173, 205, 149, 115, 114 (by column)
  - R-squared: 0.045, 0.132, 0.12, 0.222, 0.273, 0.326, 0.157, 0.161, 0.088, 0.276, 0.367

### EMBI Spreads and Liquidity: Emerging Economies (Table 5)
- Dependent variable: the first difference of the natural logarithm of the EMBI spread. Estimation: fixed effects panel regressions with standard errors clustered by country. Robust p-values in brackets. *, **, *** denote significance at 10%, 5% and 1% levels respectively.
- Coefficients (columns (1) all, (2) Latin America, (3) Asia, (4) Europe):
  - L.dlembi:
    - (1) 0.126* [0.0847]
    - (2) 0.132 [0.233]
    - (3) 0.109** [0.0118]
    - (4) 0.107 [0.214]
  - Stock Liquidity:
    - (1) 0.0357* [0.0567]
    - (2) 0.0539* [0.0936]
    - (3) 0.0620* [0.0524]
    - (4) -0.0345 [0.344]
  - US Bond Liquidity:
    - (1) -1.044** [0.0249]
    - (2) -1.781 [0.135]
    - (3) -0.564** [0.0180]
    - (4) -0.552 [0.219]
  - US Stock Liquidity:
    - (1) 0.176*** [0.000800]
    - (2) 0.243** [0.0151]
    - (3) 0.133** [0.0233]
    - (4) 0.139 [0.154]
  - US Money Market Liquidity:
    - (1) -62.2 [0.126]
    - (2) -58.81 [0.582]
    - (3) -63.79 [0.108]
    - (4) -57.95 [0.449]
  - Constant:
    - (1) -0.0142 [0.175]
    - (2) -0.0112 [0.583]
    - (3) -0.0321*** [0.00532]
    - (4) 0.0203 [0.514]
- Sample and fit:
  - Observations: 1028, 394, 416, 218 (by column)
  - # of countries: 16, 6, 6, 4 (by column)
  - R-squared: 0.055, 0.078, 0.047, 0.03

### Price-Earnings (PE) Ratios and Liquidity (Table 6)
- Dependent variable: the first difference of price-earning (PE) ratio. Estimation: fixed effects panel regressions, with standard errors clustered by country. Robust p-values in brackets. *, **, *** denote significance at 10%, 5% and 1% levels respectively.
- Coefficients across columns (1) All, (2) G-7 countries, (3) All Developed, (4) Emerging Countries, (5) Emerging Latin America, (6) Emerging Asia, (7) Emerging Europe:
  - L.pe:
    - (1) -0.296*** [0.000]
    - (2) -0.0472*** [0.001]
    - (3) -0.0537*** [0.000]
    - (4) -0.334*** [0.000]
    - (5) -0.363*** [0.000]
    - (6) -0.128*** [0.000]
    - (7) -0.109** [0.024]
  - World PE ratio:
    - (1) 0.168*** [0.004]
    - (2) 0.0724** [0.014]
    - (3) 0.0692*** [0.000]
    - (4) -0.0493 [0.452]
    - (5) -0.156 [0.238]
    - (6) 0.0553* [0.065]
    - (7) 0.0259 [0.564]
  - Stock Market Liquidity:
    - (1) 1.242 [0.518]
    - (2) -0.866* [0.055]
    - (3) -0.797*** [0.004]
    - (4) 3.946 [0.245]
    - (5) 12.55 [0.239]
    - (6) -0.0692 [0.894]
    - (7) -0.117 [0.804]
  - US Stock Market Liquidity:
    - (1) -3.624** [0.025]
    - (2) -1.411*** [0.003]
    - (3) -1.421*** [0.000]
    - (4) -5.204** [0.018]
    - (5) -7.579* [0.081]
    - (6) -2.831** [0.013]
    - (7) -0.856* [0.098]
  - Constant:
    - (1) 2.180* [0.063]
    - (2) 0.17 [0.612]
    - (3) 0.253 [0.361]
    - (4) 5.855*** [0.000]
    - (5) 6.726** [0.028]
    - (6) 1.542** [0.016]
    - (7) 1.38 [0.125]
- Sample and fit:
  - Observations: 6931, 2241, 3715, 3216, 1060, 1384, 772 (by column)
  - # of countries: 30, 7, 12, 18, 6, 7, 5 (by column)
  - R-squared: 0.175, 0.064, 0.062, 0.196, 0.212, 0.077, 0.058

*Source: IMF working paper tables and regression results provided in the content unit.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2009/_wp0952.pdf_