## 1. Private Sector Financing, 2002-2007

## Source details

**Canonical URL:** [1. Private Sector Financing, 2002-2007](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11213.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11213.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11213.pdf.json)

---

### I. Introduction and research question
- Sample: 84 economic crises in 17 advanced economies covering 1960 to 2007.
- Research hypothesis: the financial structure of an economy (bank-based vs. market-based) matters for its ability to recover from economic crises.
- Recoveries dated using the April 2009 World Economic Outlook (WEO) database (IMF, 2009) and the BBQ methodology; emphasis on “durable” recoveries with minimum duration of four quarters (6 recovery episodes of only two or three quarters excluded).

### II. Main empirical findings
- Market-based economies recover faster than bank-based ones:
  - Gap estimated between 0.8 and 1.4 percentage points two years into the recovery.
  - Comparing strongly market-based to strongly bank-based economies: positive cumulative growth gap of up to 2.7 percent two years after a trough.
  - Growth differential increases over time and is greater 8 quarters into a recovery than after 4 quarters.
- Alternative reported point estimates:
  - Difference estimated at 0.7 percentage points of cumulative growth 4 quarters after the trough and 1.4 percent 8 quarters after the trough.
  - Without the United Kingdom, coefficients increase to 0.9 and 1.6 percentage points.
  - Strongly market-based vs. strongly bank-based: 1.4 percentage points after 4 quarters, widening to 2.7 percentage points after 8 quarters.
  - Excluding the United Kingdom, cumulative growth gap can reach up to 3.6 percentage points after 8 quarters.
- Evidence of a threshold effect: the dummy for strongly market-based economies is more significant than a linear market-share measure, suggesting marginal gains increase sharply beyond a threshold.
- Financial crises negatively impact recoveries:
  - Financial Crisis dummy: -2.07*** (4 Quarters), -2.79*** (8 Quarters) in baseline.
  - Results similar excluding the United Kingdom: -2.06*** (4 Quarters), -2.76*** (8 Quarters).

### III. Stylized quantitative facts from the sample
- Countries in sample: Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Italy, Japan, Netherlands, Norway, Portugal, Spain, Sweden, the United Kingdom, the United States.
- Distribution of recoveries: 35 episodes for market-based economies and 49 episodes for bank-based economies.
- Four quarters into recoveries:
  - Cumulative GDP growth gap is 0.8 percent between market-based and bank-based economies.
  - Market-based economies reach an average cumulative growth of 3.9 percent four quarters into recovery.
- Selected averages (Recoveries and cumulative growth, 1960-2007):
  - Market-based countries: Number of Recoveries: 35; 4 Q to Trough: 3.9; 8 Q to Trough: 7.3.
  - Market-based excluding UK: Number of Recoveries: 31; 4 Q to Trough: 4.2; 8 Q to Trough: 7.5.
  - Bank-based countries: Number of Recoveries: 49; 4 Q to Trough: 3.1; 8 Q to Trough: 5.4.
  - Strongly Market-based: Number of Recoveries: 19; 4 Q to Trough: 4.2; 8 Q to Trough: 7.9.
  - Intermediate countries: Number of Recoveries: 44; 4 Q to Trough: 3.5; 8 Q to Trough: 6.1.
  - Strongly Bank-based: Number of Recoveries: 21; 4 Q to Trough: 2.7; 8 Q to Trough: 4.9.

### IV. Private sector financing (2002-2007) — country shares (Liabilities to Markets / Liabilities to Banks, average in percent)
- Australia: Markets 0.68 Banks 0.32
- Austria: Markets 0.34 Banks 0.66
- Belgium: Markets 0.26 Banks 0.74
- Canada: Markets 0.64 Banks 0.36
- Denmark: Markets 0.56 Banks 0.44
- Finland: Markets 0.58 Banks 0.42
- France: Markets 0.53 Banks 0.47
- Germany: Markets 0.39 Banks 0.61
- Italy: Markets 0.42 Banks 0.58
- Japan: Markets 0.44 Banks 0.56
- Netherlands: Markets 0.48 Banks 0.52
- Norway: Markets 0.43 Banks 0.57
- Portugal: Markets 0.29 Banks 0.71
- Spain: Markets 0.30 Banks 0.70
- Sweden: Markets 0.45 Banks 0.55
- United Kingdom: Markets 0.62 Banks 0.38
- United States: Markets 0.73 Banks 0.27
- Classification summary:
  - Market-based: Australia, Canada, Denmark, Finland, France, the United Kingdom, the United States.
  - Bank-based: Austria, Belgium, Germany, Italy, Japan, Netherlands, Norway, Portugal, Spain, Sweden.
  - Strongly market-based: United States, Australia, Canada, United Kingdom.
  - Strongly bank-based: Belgium, Portugal, Spain, Austria.

### V. Empirical evidence and key regression results (selected coefficients)
- Baseline regression highlights (4 Quarters / 8 Quarters):
  - Market-based Dummy: 0.66 (4 Quarters), 1.38** (8 Quarters).
  - MB vs BB Linear Measure: 2.79* (4 Quarters), 5.68** (8 Quarters).
  - Strongly MB vs Strongly BB: 1.39** (4 Quarters), 2.72*** (8 Quarters).
  - Financial Crisis dummy: -2.07*** (4 Quarters), -2.79*** (8 Quarters).
  - Recession Amplitude: -0.36*** (4 Quarters), -0.55*** (8 Quarters).
  - Government Consumption (proxy for contra-cyclical fiscal policy): 0.16* (4 Quarters), 0.36** (8 Quarters).
- Results excluding the United Kingdom:
  - Market-based Dummy: 0.93** (4 Quarters), 1.59** (8 Quarters).
  - MB vs BB Linear Measure: 3.58** (4 Quarters), 6.26*** (8 Quarters).
  - Strongly MB vs Strongly BB: 1.82*** (4 Quarters), 3.06*** (8 Quarters).
  - Financial Crisis dummy: -2.06*** (4 Quarters), -2.76*** (8 Quarters).
  - Recession Amplitude: -0.41*** (4 Quarters), -0.64*** (8 Quarters).
  - Government Consumption: 0.15* (4 Quarters), 0.36** (8 Quarters).

### VI. Robustness checks and role of real-sector flexibility
- Additional controls: employment protection (labor market flexibility), goods market regulation (product market flexibility), synchronized recession dummy, interaction between financial crisis and market-structure measure.
- Key takeaways:
  - Synchronized crises result in more protracted recoveries.
  - Labor market flexibility allows for faster recoveries.
  - Controlling for labor flexibility reduces the comparative advantage of market-based economies in recoveries by roughly 40 percent at both 4 and 8 quarters after the trough.
  - Interaction coefficients (FC*Market-based Measure) are negative but not significant in baseline extended specifications, suggesting the market-based advantage may be smaller during financial crises.

### VII. Interpretation of control variables
- Financial Crisis dummy: negative and significant — recoveries associated with financial crises are more sluggish.
- Recession Amplitude: negative and significant — larger cumulative losses from peak to trough are associated with more sluggish recoveries.
- Government Consumption (contra-cyclical fiscal policy): positive — tends to increase speed of recoveries.
- Employment protection and product market regulation: suggestive roles for real-sector flexibility; coefficients are sometimes not significant.

### VIII. Data, classification choices, and caveats
- Classification criterion: country classified as market-based when funding to the non-financial private sector from market sources exceeds funding from banks, using OECD National Financial Accounts data averaged over 2002-2007.
- Measure details:
  - Share of market financing = Securities other than Shares, except Financial Derivatives (F33) and Quoted Shares (F511).
  - Liabilities to the Banks = Loans (F4). Financial Sector Excluded. Non consolidated data.
- Missing-data treatments:
  - F511 not available for Canada, Italy and Japan. F511 estimated based on F51 and the assumption that F511/F51 is roughly constant in the sample (proxy used: F51 ×0.4 for missing F511).
  - No Non-Consolidated data for Australia. Consolidated data are used for Australia instead.
- Caveats:
  - Classification treated as structural and constant over time due to data limitations—may be strong for borderline countries (e.g., France, Finland, Denmark).
  - Instrument-based definition may overestimate market-based financing where bond financing is used as covert bank lending.
  - Sensitivity to data sources; a “strongly market-based” and “strongly bank-based” subgrouping is used to maximize within-group homogeneity.

### IX. Policy implications and recommendations
- Short term:
  - Prioritize dealing with weak bank balance sheets through recapitalization, restructuring, and resolving impaired assets to support recovery, particularly in bank-based economies.
- Medium term:
  - Structural policies to deepen financial markets so they can effectively complement banking sectors are useful; policies that would stifle post-crisis development of financial markets could be misguided.
- Stability concerns:
  - Development of financial markets must be accompanied by measures to enhance the stability of financial markets because financial crises are more costly in terms of slow recoveries.
- Broader reforms:
  - Policies must go beyond financial markets and address rigidities more broadly in the real economy (e.g., employment and product market flexibility) to facilitate recoveries.
- Open research agenda:
  - Additional research needed to identify more accurately the channels through which market-based economies build their comparative advantage.

*Source: _wp11213 - 1. Private Sector Financing, 2002-2007 (excerpts).*

### 1. Private Sector Financing, 2002-2007 .................................................................................

### 1. Private Sector Financing, 2002-2007

### I. Introduction and research question
- Study uses a sample of 84 economic crises in 17 advanced economies covering 1960 to 2007.
- Research hypothesis: the financial structure of an economy (bank-based vs. market-based) matters for its ability to recover from economic crises.
- Recoveries are dated using the April 2009 World Economic Outlook (WEO) database (IMF, 2009) and the BBQ methodology; emphasis is on “durable” recoveries with minimum duration of four quarters (6 recovery episodes of only two or three quarters excluded).

### II. Main empirical findings
- Within the sample of advanced economies, market-based economies recover faster than bank-based ones.
  - After controlling for a set of variables, the gap is estimated between 0.8 and 1.4 percentage points two years into the recovery.
  - Comparing strongly market-based to strongly bank-based economies: the former experience a positive cumulative growth gap of up to 2.7 percent, compared to the latter, two years after a trough.
  - The growth differential increases over time and is greater 8 quarters into a recovery than after 4 quarters.
- When employment and product market flexibility are taken into account, the comparative advantage of market-based economies in recoveries becomes less significant.
- Financial crises negatively impact the ability of market-based economies to recover compared to bank-based economies.

### III. Stylized quantitative facts from the sample
- Sample: 84 recoveries for 1960-2007 across 17 countries (Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Italy, Japan, Netherlands, Norway, Portugal, Spain, Sweden, the United Kingdom, the United States).
- Distribution of recoveries: 35 episodes for market-based economies and 49 episodes for bank-based economies.
- Four quarters into recoveries, the cumulative GDP growth gap is 0.8 percent between market-based and bank-based economies:
  - Market-based economies reach an average cumulative growth of 3.9 percent four quarters into recovery.
  - (Bank-based economies achieve the complementary figure implied by the 0.8 percent gap as reported in the source.)
- Classification of countries (based on average 2002-2007 private sector financing: liabilities to markets vs. liabilities to banks):
  - Market-based: Australia, Canada, Denmark, Finland, France, the United Kingdom, the United States.
  - Bank-based: Austria, Belgium, Germany, Italy, Japan, Netherlands, Norway, Portugal, Spain, Sweden.
  - Strongly market-based: United States, Australia, Canada, United Kingdom.
  - Strongly bank-based: Belgium, Portugal, Spain, Austria.
- Table 1 sample entries (Private Sector Financing, 2002-2007; liabilities to Markets / liabilities to Banks, average in percent):
  - Australia: Markets 0.68 Banks 0.32
  - Austria: Markets 0.34 Banks 0.66
  - Belgium: Markets 0.26 Banks 0.74
  - Canada: Markets 0.64 Banks 0.36
  - Denmark: Markets 0.56 Banks 0.44
  - Finland: Markets 0.58 Banks 0.42
  - France: Markets 0.53 Banks 0.47
  - Germany: Markets 0.39 Banks 0.61
  - Italy: Markets 0.42 Banks 0.58
  - Japan: Markets 0.44 Banks 0.56
  - Netherlands: Markets 0.48 Banks 0.52
  - Norway: Markets 0.43 Banks 0.57
  - Portugal: Markets 0.29 Banks 0.71
  - Spain: Markets 0.30 Banks 0.70
  - Sweden: Markets 0.45 Banks 0.55
  - United Kingdom: Markets 0.62 Banks 0.38
  - United States: Markets 0.73 Banks 0.27
- Data notes from Table 1:
  - Private Sector financing, Liabilities to the market: Securities other Shares, except Financial Derivatives (F33) and Quoted Shares (F511), Liabilities to the Banks: Loans (F4). Financial Sector Excluded. Non consolidated data.
  - F511 not available for Canada, Italy and Japan. F511 is estimated based on F51 (Shares and other equities, except mutual fund shares) and the assumption that F511/F51 is roughly constant in the sample.
  - No Non-Consolidated data for Australia. Consolidated data are used for Australia instead.
- Figure 1 summary (Private Sector Financing, 2002-2007): visual distribution of liabilities to Banks vs. liabilities to Markets across the 17 countries (percent shares preserved in Table 1).

### IV. Literature context and theoretical channels
- Prior evidence:
  - Financial crises are associated with slower and more protracted recoveries (Reinhart and Rogoff, 2008, 2010). April 2009 WEO: real growth one year from trough after a financial crisis is approximately 2½ percentage points lower than after other types of crises.
  - Deeper recessions tend to induce faster recoveries (“bounce-back” or plucking model).
  - Synchronized crises prolong recoveries; IMF (2009) estimates about 50 percent more time to recover when recessions are synchronized.
  - Evidence is mixed on the role of global integration (trade openness, capital account liberalization, exchange rate liberalization).
  - Countercyclical fiscal policy: a one-standard deviation increase in government spending during a recession results in a 0.7 percentage point increase in growth rate one year after trough (effect depends on initial public debt). Monetary policy: a one-standard deviation reduction of interest rates beyond Taylor-rule-implied levels yields a 0.4 percentage point gain.
- Theoretical motivations for market-based advantage in recoveries:
  - Financial markets can finance a wider range of projects/firms and may be better at allocating funds when diversity of opinion is high (Allen and Gale).
  - Banks may extract rents from firms due to insider information, reducing firms’ incentives to invest (Rajan, 1992).
  - Market-based systems may offer more sophisticated, flexible, tailor-made risk management solutions (Levine, 2000).
- Theoretical motivations for bank-based advantage:
  - Banks can mitigate asymmetric information and agency problems through long-term relationships and monitoring; scale effects and bank monitoring can support large firms (Allen and Carletti, 2008).

### V. Data, classification choices, and caveats
- Classification criterion: a country is classified as market-based when funding to the non-financial private sector from market sources exceeds funding from banks, using OECD National Financial Accounts data averaged over 2002-2007.
- Advantages of this measure:
  - Focuses on financing provision to the real economy (excludes public sector and private financial sector).
  - Includes both bond market and equity market indicators (book values used to avoid market-capitalization price effects).
- Caveats and limitations:
  - Classification treated as structural and constant over time due to data limitations—may be strong for borderline countries (e.g., France, Finland, Denmark).
  - Instrument-based definition may overestimate market-based financing where bond financing is used as covert bank lending (e.g., private placements, tax-induced distortions).
  - Sensitivity of classification to data sources; therefore, a “strongly market-based” and “strongly bank-based” subgrouping is identified to maximize within-group homogeneity.

### VI. Policy implications highlighted by the paper
- Short term:
  - Because financial structure matters, policies to reduce hurdles to recoveries in bank-based economies—such as dealing with weak bank balance sheets through recapitalization, restructuring, and resolving impaired assets—would support recovery, particularly in continental European countries.
- Medium term:
  - Structural policies to deepen financial markets so they can effectively complement banking sectors are useful; policies that would stifle post-crisis development of financial markets could be misguided.
- Stability concerns:
  - Development of financial markets must be accompanied by measures to enhance the stability of financial markets because financial crises are more costly in terms of slow recoveries.
- Broader reforms:
  - Policies must go beyond financial markets and address rigidities more broadly in the real economy (e.g., employment and product market flexibility) to facilitate recoveries.

*Source: _wp11213 - 1. Private Sector Financing, 2002-2007 (excerpts).*

### 3.1 percent growth.

### _wp11213 - 3.1 percent growth.

### Main findings
- Market-based economies recover faster than bank-based economies:
  - Difference estimated at 0.7 percentage points of cumulative growth 4 quarters after the trough and 1.4 percent 8 quarters after the trough.
  - Without the United Kingdom, coefficients increase to 0.9 and 1.6 percentage points.
- Strongly market-based vs. strongly bank-based:
  - Positive growth differential estimated at 1.4 percentage points after 4 quarters, widening to 2.7 percentage points after 8 quarters.
  - When excluding the United Kingdom, the cumulative growth gap can reach up to 3.6 percentage points after 8 quarters.
- The comparative advantage of market-based economies is durable over the 8 quarters (does not erode on the 5th to 8th quarters).
- Evidence of a threshold effect:
  - A higher share of market financing is associated with faster recoveries, but the linear measure is less statistically significant than a dummy for strongly market-based economies, suggesting marginal gains increase sharply beyond a threshold.

### Empirical evidence and key regression results
- Sample: 84 recovery episodes (35 for market-based economies and 49 for bank-based economies) over the 1960-2007 period.
- Baseline regression outcome highlights:
  - Market-based Dummy: 0.66 (4 Quarters), 1.38** (8 Quarters).
  - MB vs BB Linear Measure: 2.79* (4 Quarters), 5.68** (8 Quarters).
  - Strongly MB vs Strongly BB: 1.39** (4 Quarters), 2.72*** (8 Quarters).
  - Financial Crisis dummy: -2.07*** (4 Quarters), -2.79*** (8 Quarters).
  - Recession Amplitude: -0.36*** (4 Quarters), -0.55*** (8 Quarters).
  - Government Consumption (proxy for contra-cyclical fiscal policy): 0.16* (4 Quarters), 0.36** (8 Quarters).
- Results excluding the United Kingdom (Table 4bis):
  - Market-based Dummy: 0.93** (4 Quarters), 1.59** (8 Quarters).
  - MB vs BB Linear Measure: 3.58** (4 Quarters), 6.26*** (8 Quarters).
  - Strongly MB vs Strongly BB: 1.82*** (4 Quarters), 3.06*** (8 Quarters).
  - Financial Crisis dummy: -2.06*** (4 Quarters), -2.76*** (8 Quarters).
  - Recession Amplitude: -0.41*** (4 Quarters), -0.64*** (8 Quarters).
  - Government Consumption: 0.15* (4 Quarters), 0.36** (8 Quarters).

### Recoveries and cumulative growth (1960-2007) — selected averages from Tables
- Market-based countries (Total/Average):
  - Number of Recoveries: 35
  - 4 Q to Trough: 3.9
  - 8 Q to Trough: 7.3
- Market-based excluding UK (Total/Average):
  - Number of Recoveries: 31
  - 4 Q to Trough: 4.2
  - 8 Q to Trough: 7.5
- Bank-based countries (Total/Average):
  - Number of Recoveries: 49
  - 4 Q to Trough: 3.1
  - 8 Q to Trough: 5.4
- Strongly Market-based (Total/Average):
  - Number of Recoveries: 19
  - 4 Q to Trough: 4.2
  - 8 Q to Trough: 7.9
- Intermediate countries (Total/Average):
  - Number of Recoveries: 44
  - 4 Q to Trough: 3.5
  - 8 Q to Trough: 6.1
- Strongly Bank-based (Total/Average):
  - Number of Recoveries: 21
  - 4 Q to Trough: 2.7
  - 8 Q to Trough: 4.9

### Robustness checks and role of real-sector flexibility
- Additional controls introduced: employment protection (labor market flexibility), goods market regulation (product market flexibility), synchronized recession dummy, and interaction between financial crisis and market-structure measure.
- Key robustness takeaways:
  - Synchronized crises result in more protracted recoveries.
  - Labor market flexibility allows for faster recoveries.
  - When employment and product market flexibility are included, the comparative advantage of market-based economies becomes less statistically significant:
    - Controlling for labor flexibility reduces the comparative advantage of market-based economies in recoveries by roughly 40 percent at both 4 and 8 quarters after the trough.
  - Interaction coefficients (FC*Market-based Measure) are negative but not significant in baseline extended specifications, suggesting the market-based advantage may be smaller during financial crises.

### Interpretation of control variables
- Financial Crisis dummy: negative and significant — recoveries associated with financial crises are more sluggish.
- Recession Amplitude: negative and significant — larger cumulative losses from peak to trough are associated with more sluggish recoveries (contradicting some past studies).
- Government Consumption (contra-cyclical fiscal policy): positive — tends to increase speed of recoveries.
- Employment protection and product market regulation: suggestive roles for real-sector flexibility; evidence is less clear and some coefficients are not significant.

### Policy implications and conclusions
- Among advanced countries, market-based economies recover significantly faster than bank-based economies.
- Market-based economies are associated with a gain in cumulative economic growth of 0.4 to 0.7 percentage point on average after a year of recovery, and of 0.8 to [text truncated in source].
- Policy guidance:
  - Development of financial markets should be accompanied by measures to enhance financial stability, because financial crises lead to slower recoveries.
  - Focusing solely on financial-market structure is shortsighted; macroeconomic and real-sector conditions (especially labor-market flexibility) matter for recovery dynamics.

*Source: _wp11213 - 3.1 percent growth.*

### 1.4 percentage points two years into the recovery.

### _wp11213 - 1.4 percentage points two years into the recovery.

### Major empirical findings
- Recovery pace: "1.4 percentage points two years into the recovery."
- Financial-crisis recoveries:
  - Financial crises tend to be more protracted and followed by slower recoveries than other recessions, especially when associated with housing busts.
  - Synchronization of crises imposes a further drag on recoveries.
  - Recoveries vary substantially across countries, producing a "multi-speed recovery."
- Financial system differences:
  - Differences in financial systems are a significant factor behind country divergences in recovery performance.
  - Market-based economies display comparatively faster recoveries in part because market financing complements bank financing.

### Policy implications and recommendations
- Bank repair priority:
  - Banking sector repair is paramount to avoid slow and protracted recoveries, especially in bank-based economies due to the slow recovery in bank lending after crises.
- Market financing:
  - Market financing needs to be well supervised, but it provides a useful complement to bank financing.
- Open research agenda:
  - Additional research is needed to identify more accurately the channels through which market-based economies build their comparative advantage.

### Appendix — Data considerations (methods and measurements)
- Business cycle dating and sample selection:
  - Peaks and troughs dated using the April 2009 World Economic Outlook database.
  - To focus on “durable” recoveries, a phase must last at least four quarters (instead of two in IMF WEO (2009)); as a result, 6 recovery episodes were excluded.
  - Identification of financial crises based on IMF WEO (2009).
  - Synchronized recessions rely on IMF WEO (2009); 2001 was added as a synchronized episode because 12 out of the 17 countries entered into a recession.
- Measure of Market-Based / Bank-Based:
  - Built using the non consolidated Financial National Account database.
  - Share of market financing = sum of “Securities other than Shares, except Financial Derivatives” (F33) and “Quoted Shares” (F511).
  - Share of market lending = “Loans - Financial Sector Excluded” (F4).
- Missing-data treatments and proxies:
  - “Quoted Shares” (F511) was missing for Canada, Italy and Japan. Proxy used: “Shares and Equity, except Mutual Fund” (F51) ×0.4. Rationale: for the 14 countries with both F511 and F51, F511 accounted for roughly 40 percentage points of F51.
  - For Australia, non consolidated data were missing; consolidated data from the Financial National Account database were used instead.
- Control variables and external datasets:
  - “Government Spending” from IMF WEO (2009).
  - “Strictness of Employment Protection – Overall, version 3” from OECD.
  - “Product Market Regulation” from OECD.

*Source: IMF working paper content in _wp11213 - 1.4 percentage points two years into the recovery.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2011/_wp11213.pdf_