## _wp08274 — Executive Summary and Section VIII summary

## Source details

**Canonical URL:** [_wp08274 — Executive Summary and Section VIII summary](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08274.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08274.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08274.pdf.json)

---

### Scope, data and methodology
- Sample: 21 OECD countries, 1960:1-2007:4; countries listed in source.
- Events identified with Harding and Pagan (2002a) BBQ algorithm on log-levels; main dating series = output (GDP).
- Minimum duration censoring:
  - Minimum duration of a complete cycle: 5 quarters.
  - Minimum duration of each phase: 2 quarters.
  - Exception for asset prices: contraction-phase minimum 2 quarters ignored if quarterly decline exceeds 20 percent.
- Event definitions (top-quartile thresholds unless otherwise noted):
  - Severe recession: peak-to-trough output decline below -3.2 percent (top quartile).
  - Credit crunch: peak-to-trough contraction in credit in top quartile (credit crunch threshold noted as >9.5 percent decline for crunch definition elsewhere).
  - House price bust: peak-to-trough decline in house prices in top quartile (house bust threshold noted as >14.3 percent decline).
  - Equity price bust: peak-to-trough decline in equity prices in top quartile (equity bust threshold noted as >38.7 percent decline).
- Policy proxies:
  - Fiscal: government consumption (OECD Analytical Database), real (deflated by CPI).
  - Monetary: short-term interest rates (IFS, Haver, Datastream), nominal and real.

### Event counts and basic tallies
- Recessions: 122.
- Credit contraction episodes: 112 (28 crunches).
- House price decline episodes: 114 (28 busts).
- Equity price decline episodes: 234 (58 busts).
- Overlaps with recessions:
  - Recessions associated with credit crunches: 18.
  - Recessions associated with house price busts: 34.
  - Recessions associated with equity price busts: 45.
- Recessions associated with crunches and busts: 76; Recessions associated with none: 46.

### Typical recession features (stylized facts)
- Duration and amplitude:
  - Typical recession lasts almost 4 quarters (median duration 3.00 quarters; mean duration 3.64).
  - Typical peak-to-trough output drop: roughly 2 percent (median amplitude -1.87; mean amplitude -2.63).
  - Typical cumulative loss: around 3 percent (median cumulative loss -3.04; mean cumulative loss -6.40).
  - Severe recessions (30 episodes) median decline about 5 percent and median cumulative loss about 10 percent.
  - Depression definition: peak-to-trough decline exceeds 10 percent; 5 depressions documented.
- Frequency:
  - Typical country experienced about 5 recessions over 1960-2007 (Median Number of Recessions: 5.00; Mean Number: 5.81).
  - Typical fraction of time spent in recession: around 10–11 percent (Median Proportion of time in recession: 0.11).
- Time pattern:
  - Typical amplitude fell from 2.6 percent in 1973-1985 to 1.4 percent in 1986-2007 (Great Moderation pattern noted).

### Macroeconomic and financial dynamics around recessions
- General patterns:
  - Macroeconomic and financial variables are generally procyclical during recessions: sharp declines in (residential) investment, industrial production, imports, housing and equity prices; modest declines in consumption and exports; some decrease in employment rates.
  - Short-term interest rates and fiscal expenditures often behave countercyclically.
- Magnitudes and timings:
  - Median quarterly decline in output during recessions: around -0.5 percent; typical recession leads to roughly 1.5 percent decline in output per quarter compared with expansions in one quoted passage.
  - Consumption: average contraction much smaller than in output; in severe recessions consumption typically drops by more than 1 percent.
  - Investment: residential and total investment decline by double digits in severe recessions; in other recessions investment drops about 4 percent.
  - Industrial production: decline larger than output decline and tracks investment closely.
  - Trade: imports fall much more than exports; imports can fall close to 10 percent in severe recessions.
  - Unemployment: increases in 90 percent of recessions; rise typically three times larger in severe recessions.
  - Inflation: typically drops slightly in 60 percent of recessions; some severe recessions show stagflation.
- Financial series behavior:
  - Credit typically continues to grow during recessions but at about 1 percent; credit growth slows by 2–3 percentage points before a recession and by another 2 percentage points over the recession period, often not returning to pre-recession growth for at least three years.
  - House prices fall in around 55 percent of recessions; equity prices fall in about 60 percent.
  - Median equity price decrease in severe recessions: 16 percent.

### Duration and amplitude of financial episodes (exact summaries)
- Aggregate durations and amplitudes (from Table 5 summary):
  - Credit Contractions: Mean duration 5.65; Median amplitude -4.22.
  - Credit Crunches: Mean duration 10.3***; Median amplitude -17.03***.
  - House Price Declines: Mean duration 8.47; Median amplitude -5.99.
  - House Price Busts: Mean duration 18.43***; Median amplitude -28.52***.
  - Equity Price Declines: Mean duration 6.91; Median amplitude -26.58.
  - Equity Price Busts: Mean duration 11.78***; Median amplitude -50.27***.
- Typical episode characterizations:
  - A credit crunch typically lasts two-and-a-half years and is associated with nearly a 20 percent decline in credit (alternative mean duration characterization: mean duration around 10 quarters; median amplitude -17.03).
  - A housing bust tends to persist four-and-a-half years with a 30 percent fall in real house prices (Table 5 median amplitude -28.52; mean duration 18.43).
  - An equity price bust lasts some 10 quarters with a roughly 50 percent price decline (median amplitude -50.27; mean duration 11.78).

### Coincidence, overlap and timing between financial episodes and recessions
- Overlap frequencies:
  - One out of six recessions coincides with a credit crunch.
  - One out of four recessions coincides with a house price bust.
  - One-third of recessions coincide with equity price busts.
- Leads and lags (median values, Table 6):
  - Median lead (quarters between start of crunch/bust and start of recession):
    - Credit Crunches: 4.00.
    - House Price Busts: 3.00.
    - Equity Price Busts: 5.00.
  - Median lag (quarters between end of recession and end of crunch/bust):
    - Credit Crunches: 2.00.
    - House Price Busts: 9.00.
    - Equity Price Busts: 0.00.
- Synchronization:
  - Recessions and credit contractions are highly synchronized across countries; house and equity price declines also show strong synchronization, with equity prices exhibiting the highest degree of synchronization.

### Comparative severity: recessions with financial distress vs. other recessions
- Recessions associated with credit crunches and house price busts are deeper and longer:
  - Example medians (Tables 7–9):
    - With Credit Crunches: Median Amplitude -2.19 vs. -1.82 without; Median Cumulative Loss -4.44 vs. -2.87 without.
    - With House Price Busts: Median Duration 3.20 vs. 3.00 without; Median Amplitude -2.18 vs. -1.52 without; Median Cumulative Loss -3.74 vs. -2.25 without.
    - With Equity Price Busts: Median Amplitude -1.98 vs. -1.63 without; Median Cumulative Loss -3.08 vs. -2.64 without.
  - Residential investment falls more sharply in recessions with housing busts and in those with credit crunches; unemployment rises more in recessions with housing busts.
  - Differences in total output loss between events with severe crunches/busts and those without typically amount to one percentage point in some comparisons.

### Recessions associated with oil price shocks and stagflation
- Definition: oil price "shock" = increase in top quartile (top 12.5 percent in some contexts) of price increases.
- Almost half of recessions in sample are associated with an oil price shock.
- Comparative medians (Table 11):
  - Median Amplitude: -2.05 (with oil shocks) vs. -1.78 (without).
  - Median Cumulative Loss: -3.14 (with) vs. -2.94 (without).
  - Median Inflation Rate change: 0.29 (with) vs. -0.90 (without).
- Recessions with oil price shocks are often stagflationary: larger output drops and significantly higher inflation rates; equity and house price changes not much different from other recessions.

### Policy responses (median changes from Table 12)
- General pattern: both monetary and fiscal policies tend to be countercyclical; fiscal policy appears more accommodative in severe recessions, credit crunches and asset price busts.
- Selected median changes:
  - Recessions:
    - Short-Term Nominal Interest Rate: -0.79.
    - Short-Term Real Interest Rate: -0.70.
    - Government Consumption (percent change): 1.79.
  - Credit Crunches:
    - Short-Term Nominal Interest Rate: -1.50.
    - Short-Term Real Interest Rate: -0.09.
    - Government Consumption: 6.33***.
  - House Price Busts:
    - Short-Term Nominal Interest Rate: -3.16***.
    - Short-Term Real Interest Rate: 0.21.
    - Government Consumption: 9.07***.
  - Equity Price Busts:
    - Short-Term Nominal Interest Rate: 0.28.
    - Short-Term Real Interest Rate: -0.57.
    - Government Consumption: 7.72***.
- Interpretation:
  - Government consumption rises significantly more in episodes with credit crunches, house price and equity busts—consistent with more aggressive countercyclical fiscal policy when credit frictions present.
  - During house price busts nominal interest rate declines are statistically significantly larger than in non-bust episodes.
  - In stagflationary recessions nominal short-term interest rates do not fall as they do in other recessions (real short-term interest rate drops larger in these episodes).

### Formal regressions on determinants of recession amplitude (Section VIII)
- Baseline OLS regressions (sample: 1960:1-2007:4) include financial regressors (changes in credit, housing, equity prices during recessions) and controls:
  - Cumulative output growth over two years prior to recession (initial output state).
  - Change in exports during recessions.
  - Growth rate of oil prices in two years preceding recession.
  - Changes in government expenditures.
  - Changes in short-term real interest rates.
  - Great Moderation dummy (one after 1986:2, zero otherwise).
  - Crisis dummy (banking or currency crisis during or in year prior to recession).
- Key empirical findings (preserved sign and significance patterns):
  - Financial variables: coefficients positive—declines in credit, house prices and equity prices associated with deeper recessions; house and equity prices initially statistically significant.
  - House prices: change in house prices tends to be the most robust financial variable associated with recession depth; in specifications including all financial variables the coefficient on housing prices remains statistically significant and positive.
  - Credit: coefficient signs vary with specification; can change sign and significance.
  - Equity prices: significance can disappear in multivariate specifications.
  - Other robust determinants:
    - State of the economy at onset (pre-recession cumulative growth) positively associated with recession depth (strong expansions followed by larger contractions).
    - Decline in exports positively correlated with recession depth and significant in almost all specifications.
    - Great Moderation dummy indicates milder recessions in most specifications.
    - Recession amplitude positively associated with its duration where duration included.
  - Quantile regressions (Tables 14A-14B): main messages preserved—housing price changes remain significantly positively correlated with recession costs; credit coefficients often negative and significant in some quantiles; Great Moderation negative and significant in multiple quantiles.
  - Crisis dummy becomes positive and significant in some specifications, indicating financial crises can be positively associated with recession costs in certain models.

### Conclusions, policy lessons and caveats
- Conclusions:
  - Interactions between macroeconomic and financial variables play a key role in determining severity and duration of recessions.
  - Recessions associated with credit crunches and house price busts are deeper, longer-lasting and have larger declines in consumption and investment than other recessions.
  - House price declines emerge as the financial variable most robustly associated with recession depth across specifications.
  - Equity price busts are less consistently associated with real-sector outcomes than credit crunches and house price busts.
  - Recession severity is also strongly influenced by the state of the economy at recession onset and by recession duration; recessions during the Great Moderation tend to be milder.
- Policy-relevant lessons:
  - Historical episodes suggest recessions following the financial storm referenced in the source are likely to be more costly because they coincide with simultaneous credit crunches and asset price busts.
  - Severity for any country depends on pre-recession financial health of firms, banks, households, and on policy measures taken; decisive policy action can help mitigate outcomes.
- Caveats and future research:
  - Event-study design: regressions capture associations, not causal effects; cannot definitively establish direction of causation between financial variables and macro outcomes.
  - Policy proxies are coarse; timing and potency of policy interventions are not fully captured.
  - Suggested future research: identify channels (wealth, balance-sheet), use firm-level data, consider alternative activity metrics, and expand sample to emerging market economies.

*Source: _wp08274 — Executive Summary and Section VIII summary (IMF working paper excerpt).*

### Executive Summary ......................................................................................................

### Executive Summary

### Scope and Structure
- Covers the Executive Summary and the main sections of the paper, listed as:
  - I. Introduction
  - II. Database and Methodology
    - A. Database
    - B. Methodology
  - III. What Happens During Recessions?
    - A. Basic Features of Recessions: Duration and Cost
    - B. Changes in Macroeconomic and Financial Variables
    - C. Dynamics of Recessions
    - D. Synchronization of Recessions, Credit Contractions and Asset Price Declines
  - IV. What Happens During Credit Contractions and Asset Price Declines?
    - A. Episodes of Credit Contractions
    - B. Episodes of Declines in House Prices
    - C. Episodes of Declines in Equity Prices
    - D. Credit Contractions and Asset Price Declines: A Summary
  - V. What Happens During Recessions Associated with Crunches and Busts?
    - A. Recessions Associated with Credit Crunches
    - B. Recessions Associated with House Price Busts
    - C. Recessions Associated with Equity Price Busts
    - D. Recessions Associated with Crunches and Busts: A Summary
  - VI. Recessions Associated with Increases in Oil Prices
  - VII. Policy Responses During Recessions, Crunches and Busts
  - VIII. Recession Outcomes and Financial Factors
  - IX. Conclusion
    - A. A Summary
    - B. Lessons for Today
    - C. Caveats and Future Research

### Major Themes Covered
- Recessions: duration, amplitude, and cost (see Section III.A)
- Macroeconomic and financial variable changes during recessions (see Section III.B)
- Dynamics and synchronization of recessions, credit contractions, and asset price declines (see Sections III.C–III.D and IV.D)
- Episodes and characterization of:
  - Credit contractions (Section IV.A)
  - House price declines (Section IV.B)
  - Equity price declines (Section IV.C)
- Recessions associated specifically with credit crunches, house price busts, and equity price busts (Section V)
- Role of oil price increases in recessions (Section VI)
- Policy responses during recessions, crunches, and busts (Section VII)
- Links between financial factors and recession outcomes (Section VIII)
- Summary, lessons, and caveats for future research (Section IX)

### Empirical Assets (Figures and Tables)
- Figures listed with their titles and numbering:
  - Figure A. What Happens During Recessions: Crunches and Busts?
  - 1. Associations Between Recessions, Crunches and Busts
  - 2. Recessions: Duration and Amplitude
  - 3. Recessions in OECD Countries
  - 4. Synchronization of Recessions
  - 5. Synchronization of Credit Contractions and Asset Price Declines
  - 6. Credit Crunches in OECD Countries
  - 7. House Price Busts in OECD Countries
  - 8. Equity Price Busts in OECD Countries
- Tables listed with numbering and titles:
  - 1A. Recessions: Summary Statistics
  - 1B. Recessions: Summary Statistics
  - 2A. Credit Contractions: Basic Statistics
  - 2B. Credit Contractions: Basic Statistics
  - 3A. House Price Declines: Basic Statistics
  - 3B. House Price Declines: Basic Statistics
  - 4A. Equity Price Declines: Basic Statistics
  - 4B. Equity Price Declines: Basic Statistics
  - 5. Credit Contractons and Asset Price Declines: Summary Statistics
  - 6. Leads and Lags: Recessions, Crunches and Busts
  - 7. Recessions Associated with Credit Crunches
  - 8. Recessions Associated with House Price Busts
  - 9. Recessions Associated with Equity Price Busts
  - 10. Recessions Associated with Crunches and Busts: Summary Statistics
  - 11. Recessions Associated with Oil Price Shocks
  - 12. Changes in Poicy Variables
  - 13A. Cost of Recessions
  - 13B. Cost of Recessions
  - 14A. Cost of Recessions
  - 14B. Cost of Recessions

### Appendix and Supporting Materials
- Appendix: Database (listed)
- Page and figure/table pagination indicated in the source's front matter (pages referenced include "4", "6", "7", "10", "14", "16", "17", "18", "20", "22", "23", "24", "25", "26", "27", "28", "29", "30", "34", "35", "36", "41", "43"–"75" as appearing in the table of contents and lists)

*Source: _wp08274 - Executive Summary (IMF PDF)._

### Executive Summary

### Executive Summary

### Overview of the crisis and research questions
- The financial turmoil that started in the United States has transformed into a severe credit crunch with substantial losses in equity markets and has spread to a number of advanced and emerging countries, becoming "the most severe global financial crisis since the Great Depression."
- Two central questions addressed:
  - How do macroeconomic and financial variables behave around recessions, credit crunches and asset (house and equity) price busts?
  - Are recessions associated with credit crunches and asset price busts different than other recessions?
- Empirical characterization covers 21 OECD countries over the 1960-2007 period.

### Data coverage and event counts
- Events identified using standard business cycle dating methods:
  - 122 recessions
  - 112 (28) credit contraction (crunch) episodes
  - 114 (28) episodes of house price declines (busts)
  - 234 (58) episodes of equity price declines (busts)

### Typical recession dynamics
- The typical recession:
  - lasts almost 4 quarters
  - is associated with an output drop of roughly 2 percent
- Severe recessions (top quartile of peak-to-trough declines):
  - median decline of about 5 percent
  - last a quarter longer than typical recessions
  - typical recessions result in a cumulative loss of around 3 percent; severe ones cost three times more
- Macroeconomic and financial variables are generally procyclical during recessions:
  - sharp declines in (residential) investment, industrial production, imports, housing and equity prices
  - modest declines in consumption and exports
  - some decrease in employment rates
  - short-term interest rates and fiscal expenditures often behave countercyclically during recessions

### Duration and severity of financial episodes vs. recessions
- Credit crunches, house price busts, and equity price busts last much longer than recessions:
  - a credit crunch episode typically lasts two-and-a-half years and is associated with nearly a 20 percent decline in credit
  - a housing bust tends to persist four-and-a-half years with a 30 percent fall in real house prices
  - an equity price bust lasts some 10 quarters and when it is over, the real value of equities drops by half
- Alternative average-duration characterization:
  - average duration of a credit crunch is around 10 quarters
  - average duration of a house price bust is 18 quarters
  - average duration of an equity price bust is 12 quarters
- Investment dynamics:
  - much larger decline in the growth rate of investment compared with consumption during recessions, credit crunches and house price busts
  - credit crunches and house price busts are accompanied with large declines in residential investment
- Equity price busts are less consistently associated with real sector outcomes than credit crunches and house price busts

### Coincidence, overlap and timing between financial episodes and recessions
- Overlap statistics:
  - in one out of six recessions there is also a credit crunch underway
  - in one out of four recessions there is a house price bust
  - equity price busts coincide with one-third of recession episodes
- Timing lags:
  - a recession, if one occurs, can start as late as four to five quarters after the onset of a credit crunch or housing bust
- Synchronization:
  - recessions and credit contractions are highly synchronized across countries; declines in house and equity prices tend to occur at the same time across countries

### Comparative severity: recessions with financial distress vs. other recessions
- Recessions associated with credit crunches and house price busts are deeper and longer-lasting than other recessions:
  - although such recessions last only three months longer (in one summary), they typically result in output losses two to three times greater than recessions without such financial stresses
  - differences in total output loss between events with severe crunches/busts and those without typically amount to one percentage point
  - duration is more than one quarter longer in the case of a housing bust
  - residential investment falls more sharply in recessions with housing busts and in those with credit crunches than in other recessions
  - unemployment rates increase notably more in recessions with housing busts
- Evidence indicates the extent of declines in house prices appears to influence the depth of recessions, even after accounting for changes in other financial variables, including credit and equity prices, and various other controls

### Formal analysis and key drivers of recession depth
- A basic regression framework is used to examine how the amplitude of a recession is associated with changes in financial variables during recessions.
- Main regression finding:
  - changes in house prices tend to be the financial variable most robustly associated with the depth of recessions
  - besides duration, the extent of decline in output is most influenced by the state of the economy at the onset of the recession

### Contributions to the literature
- The paper fills three gaps:
  - examines implications of recessions, credit crunches, house and equity price busts for a large set of macroeconomic and financial variables across many countries over a long period
  - is the first detailed, cross-country empirical analysis addressing recessions coinciding with credit crunches, house price busts and equity price busts
  - provides preliminary evidence that change in house prices during recessions is an important factor influencing recession costs

### Policy-relevant lessons and outlook
- Historical episodes suggest recessions following the current crisis will likely be more costly than other recessions because they take place alongside simultaneous credit crunches and asset price busts.
- Although the effects of the current crisis have already been felt globally, past evidence suggests its global dimensions are likely to intensify in the coming months.

*Executive Summary.*

### Section VIII presents a more formal analysis of the roles played by financial factors in

### _wp08274 - Section VIII presents a more formal analysis of the roles played by financial factors in

### Database: sample, variables, and sources
- Sample: 21 OECD countries over the period 1960:1-2007:4. The countries are Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan, Netherlands, New Zealand, Norway, Portugal, Spain, Switzerland, Sweden, the United Kingdom, and the United States.
- Macroeconomic variables analyzed: output, consumption, investment, residential investment, non-residential investment, industrial production, exports, imports, net exports, current account balance, unemployment rate, inflation rate.
- Financial variables analyzed: credit (claims on the private sector by deposit money banks, from IFS), house prices (BIS), equity prices (IFS). All financial variables deflated by the respective CPI (real terms).
- Policy variables: government consumption (OECD Analytical Database) as proxy for fiscal policy; short-term interest rates (IFS, Haver Analytics, Datastream) as proxy for monetary policy. Short-term rates considered in nominal and real terms (nominal deflated by CPI). Government consumption deflated by CPI.
- Series are quarterly, seasonally adjusted where necessary, and in constant prices.
- Detailed sources and definitions of each variable are listed in Appendix I (as stated in the source).

### Methodology: dating cycles and definitions
- Business cycle definition: "Classical" definition (Burns and Mitchell (1946); NBER (2001)) — cycles identified by changes in the level of economic activity (peaks and troughs).
- Algorithm: Harding and Pagan (2002a) BBQ algorithm (extension of BB algorithm Bry and Boschan (1971)) applied to log-levels of series to identify turning points in quarterly data.
- Minimum duration censoring rules:
  - Minimum duration of a complete cycle: 5 quarters.
  - Minimum duration of each phase: 2 quarters.
  - Exception for asset prices: contraction phase minimum 2 quarters ignored if quarterly decline exceeds 20 percent.
- Peak/trough technical conditions: (algorithm specified via discrete inequalities in the source).
- Main macro variable for dating: output (GDP). Also identify cycles in consumption, investment, credit, house prices, equity prices.

### Classification of events: severe/mild, busts, crunches, and associations
- Recession severity:
  - Severe recession: peak-to-trough output decline in the top quartile of all output drops (i.e., below -3.2 percent).
  - Mild recession: peak-to-trough output drop in the bottom quartile.
- Asset/credit extremes:
  - Equity (or house) price bust: peak-to-trough decline in top quartile of equity (or house) price declines.
  - Credit crunch: peak-to-trough contraction in credit in top quartile of all credit contractions.
- Dating rule for associations: If a recession starts at the same time or after the beginning of an ongoing credit crunch or asset price bust, the recession is considered associated with that crunch or bust (a timing association, not causal by definition).
- Identified counts:
  - 122 recessions in output (30 of which are severe).
  - 112 credit contractions (28 crunches).
  - 114 house price declines (28 busts).
  - 234 equity price declines (58 busts).
- Overlap counts between recessions and financial extremes:
  - 18 recession episodes associated with credit crunches.
  - 34 recession episodes associated with house price busts.
  - 45 recession episodes associated with equity price busts.
- Additional overlap note: Some house price busts last longer than recessions; five housing busts overlap with two recessions each, two busts overlap with three recessions each.

### Validation of dating algorithm
- U.S. comparison to NBER:
  - NBER reports 7 recessions over 1960-2007 for the United States.
  - Algorithm exact matches for 4 of the 7 NBER peak/trough dates; for the remaining peaks/troughs the algorithm is a quarter early.
  - Average duration and average amplitude of U.S. cycles from the algorithm are very close to NBER measures (example: average peak-to-trough decline in output is around -1.7 percent based on algorithm dates vs -1.4 percent based on NBER dates).

### Stylized facts on recessions: frequency, duration, and cost
- Typical country experience:
  - About 5 recessions over 1960-2007 for a typical OECD country.
  - Variation: Canada, Ireland, Japan, Norway, Sweden had only 3 recessions; Italy and Switzerland had 9; New Zealand had 12 (the most).
- Duration:
  - Typical recession lasts about 4 quarters.
  - Shortest recession: 2 quarters; Longest: 13 quarters.
  - Proportion of recessions that are short (2 quarters): roughly one-third.
  - Distribution of durations: around 35 percent are 2 quarters, 40 percent are 3-4 quarters, 25 percent are 5 quarters or more.
  - Proportion of time spent in recession (fraction of quarters in recession over full sample): typically around 10 percent.
- Amplitude and cumulative loss:
  - Median (average) peak-to-trough output decline (amplitude): 1.9 (2.7) percent.
  - Range across countries: about 1 percent (Austria, Belgium, Ireland, Spain typical) to more than 6 percent (Greece, New Zealand typical).
  - Median cumulative loss of a typical recession: about 3 percent; average cumulative loss: about 6.4 percent (distribution skewed right).
  - Correlation between duration and amplitude: 0.34 (small positive).
  - Country examples: Finland and Sweden have moderate median amplitudes but very large cumulative output losses of 23 and 16 percent respectively due to long durations.
- Severe recessions (30 episodes):
  - Typical severe recession duration: 5 quarters (more than a quarter longer than average).
  - Median decline in severe recessions: about 5 percent (almost three times that of other recessions).
  - Median cumulative loss in severe recessions: about 10 percent (five times that of other recessions).
  - Depression definition: peak-to-trough decline in output exceeds 10 percent; 5 depressions documented in the sample.
  - The 5 depressions listed in the sample (as stated in source): New Zealand (1966:4-1967:2); New Zealand (1974:3-1975:2); New Zealand (1976:4-1978:1); Greece (1973:4-1974:3); Finland (1990:1-1993:2). (Source notes: Finland was the longest with 13 quarters; deepest was New Zealand with roughly 15 percent reduction in output over 1976:4-1978:1.)
- Time pattern:
  - Recessions became shorter and milder over time, especially after the mid-1980s.
  - Typical amplitude fell from 2.6 percent in 1973-1985 to 1.4 percent in 1986-2007 (consistent with the "Great Moderation" literature).

### Macro and financial behavior during recessions
- Macroeconomic variables (peak-to-trough changes; severe vs other recessions):
  - Consumption: in severe recessions typically drops by more than 1 percent; in other recessions almost no change.
  - Investment:
    - Residential and total investment decline by double digits in severe recessions.
    - In other recessions, investment drops by about 4 percent.
  - International trade:
    - Exports drop more in severe recessions than in other recessions (statistically significant).
    - Imports fall by about six times more than exports in a typical recession.
    - Imports fall close to 10 percent in severe recessions (statistically significantly more than in other recessions).
    - Net exports and current account balance typically improve during recessions; changes not statistically significantly different across recession types.
  - Industrial production: decline tracks investment drop closely and is larger than output decline.
  - Unemployment: increases in 90 percent of recessions; rise is typically three times larger in severe recessions than in other recessions.
  - Inflation: typically drops slightly in 60 percent of recessions; no systematic difference across recession types (some severe recessions exhibit stagflation).
- Financial variables:
  - Credit: typically continues to grow during recessions, but at about 1 percent, with especially low growth in initial stages of recessions. Credit growth does not vary much between severe and other recessions.
  - House prices and equity prices: typically contract during recessions.
    - House prices: larger declines in severe recessions than in other recessions.
    - Equity prices: median equity price decreases by 16 percent in severe recessions (more than twice the decline in house prices), or about 12 percent more than in other recessions.
- Quarterly median changes:
  - Median quarterly decline in output during recessions: around -0.5 percent.
  - (Quarterly expansionary comparison truncated in source text provided.)

* _wp08274 - Section VIII presents a more formal analysis of the roles played by financial factors in_

### 0.9 percent. This suggests that a typical recession leads to roughly 1.5 percent decline in

### _wp08274 - 0.9 percent. This suggests that a typical recession leads to roughly 1.5 percent decline in

### Recession impact on output
- "0.9 percent."
- "This suggests that a typical recession leads to roughly 1.5 percent decline in output per quarter compared with the periods of expansions the countries normally enjoy."

### Consumption versus output
- "The average rate of contraction in consumption was much smaller than that in output with"

*Source: _wp08274 - 0.9 percent. This suggests that a typical recession leads to roughly 1.5 percent decline in*

### 0.03 percent per quarter, but the rate of growth during expansions was close to 0.75 percent.

### _wp08274 - 0.03 percent per quarter, but the rate of growth during expansions was close to 0.75 percent.

### C. Dynamics of Recessions
- Analysis window: year-on-year growth over a 6-year window—12 quarters before and 12 quarters after a peak.
- Presentation: median growth rate with top and bottom quartiles; bottom quartile = severe recessions, top quartile = mild recessions.
- Output:
  - Following the peak at date 0, output registers a negative annual growth rate after 3 quarters.
  - Output growth rate goes down to -1 percent at the end of the fourth quarter after the peak.
  - In severe recessions the growth rate falls to -2 percent at that time.
- Consumption:
  - Typically does not decrease year-to-year in a typical recession.
  - Falls during the first year of a severe recession.
  - Timing of consumption evolution resembles output.
- Investment and industrial production:
  - Residential investment typically declines sharply ahead of recessions.
  - Both residential and non-residential investment often register negative year-to-year changes already in the first quarter of a recession (three quarters ahead of output).
  - Investment growth rates typically stay negative for up to 6 quarters; in severe recessions recovery can take up to three years.
  - Industrial production shows early weakness and typically registers a sharp decline before a recession starts.
- Inflation and unemployment:
  - Inflation typically still rising at the onset of recessions, then declines after the recession starts.
  - Unemployment typically begins climbing a quarter ahead of recessions and continues to be elevated more than a year after the end of the recession.
- Trade:
  - Growth rates of both exports and imports slow during recessions, with imports slowing much more.
  - The growth rate of imports can decline to -7 percent in the first year of a severe recession.
  - Net exports and the current account balance improve during a typical recession; improvement in net exports is earlier and more pronounced.
- Credit and asset prices:
  - Credit declines during recessions in only around 35 percent of cases (footnote context).
  - House prices fall in around 55 percent of all recessions.
  - Equity prices register a fall in about 60 percent of recessions.
  - Credit growth slows by some 2 to 3 percentage points before a recession starts, and then by another 2 percentage points over the recession period, typically not returning to pre-recession growth rates for at least three years.
  - House and equity prices decline on a year-to-year basis in the first year of a typical recession by roughly 3 and 16 percent, respectively.
  - Equity prices often start registering positive growth after about six quarters; house prices typically decline during the two years after the end of a recession.

### D. Synchronization of Recessions, Credit Contractions and Asset Price Declines
- Synchronization measure: fraction of countries experiencing the same event at the same time.
- Historical clustering (1960–2007):
  - Recessions bunch in about four periods: mid-1970s (after first oil price shock), second oil price shock/early 1980s contractionary monetary policies, early 1990s, and early 2000s to some degree.
  - In the first three of these four periods, more than 50 percent of countries in the sample were in recession at the same time.
- Investment vs. consumption synchronization:
  - Investment declines in three-fourth of all recessions; consumption contracts in only half.
  - Fraction of countries experiencing investment contractions is much higher than that experiencing recessions; consumption contractions are much less synchronized.
- Financial synchronization:
  - Recessions are highly correlated with credit contractions and bear asset markets.
  - Credit contractions are closely associated with recessions.
  - House price declines are highly synchronized across countries despite housing being nontradable; synchronization rises during recession episodes.
  - Equity prices exhibit the highest degree of synchronization.
  - Fraction of countries experiencing bear equity markets frequently exceeds the fraction in recession.

### IV. What happens during credit contractions and asset price declines?
- Definitions (top quartile thresholds):
  - Credit crunch: peak-to-trough decline in credit exceeds 9.5 percent.
  - House price bust: decline in house price larger than 14.3 percent.
  - Equity price bust: decline in equity price larger than 38.7 percent.

A. Episodes of Credit Contractions
- Counts and country experience:
  - 112 credit contraction episodes; 28 crunch episodes.
  - Typical OECD country went through about 6 credit contractions.
  - Austria, France, Germany and Switzerland never experienced a credit crunch episode during the 1960-2007 period.
- Duration and amplitude:
  - Median (average) credit contraction episode lasts 4 (6) quarters.
  - Credit crunches typically last 8 quarters (twice as long) and are statistically significantly longer than non-crunch episodes.
  - Credit contractions usually mean some 4 percent decline in credit peak-to-trough; crunches show 17 percent decline.
- Output and labor:
  - Output growth slows early in contractions/crunches; output typically higher at the end than at the beginning of these episodes.
  - Average growth rate of output in credit crunch episodes is less than half that observed during other periods.
  - Quarterly growth rate of output typically around 0.3 percent during a credit crunch versus more than 0.8 percent during other contraction episodes.
- Investment and unemployment:
  - Credit contractions (crunches) typically accompanied by declines in residential investment of about 1 (6) percent over the contraction period.
  - Unemployment typically flat during a credit contraction, but increases significantly during a credit crunch.
- House and equity prices:
  - House prices decline significantly more during credit crunches, by some 10 percent versus 1 percent in typical non-crunch episodes.
  - Equity prices usually decline somewhat during credit contractions but actually increase over credit crunch episodes.
- Dynamics around credit crunches (year-on-year, 12 quarters before/after):
  - Output growth typically starts declining two quarters before the beginning of a credit crunch and goes down by 2 percentage points after the fifth quarter.
  - In a typical credit crunch, consumption year-on-year growth can fall to -2 percent in about five quarters in some episodes.
  - Residential investment typically starts to slow much before the crunch and actually shrinks one quarter ahead of the start.
  - Growth rates of total and residential investment typically stay negative for up to 8 quarters; recovery can take up to three years (residential even longer).
  - Inflation increases prior to crunch start then declines as activity slows; unemployment increase accelerates after crunch begins.
  - Credit dynamics: median year-on-year credit growth is 5 to 6 percent just before peak of credit expansion, then slows sharply over crunch period by more than 10 percentage points, falling to -6 percent and not returning to positive levels until 10 quarters after crunch start.
  - House prices typically fall in the first year of a credit crunch and continue to decline for at least three years after the beginning; equity prices often decline before a crunch and then frequently recover ahead of credit pickup.

B. Episodes of Declines in House Prices
- Counts and durations:
  - 114 house price decline episodes; 28 busts.
  - Typical country experienced around 6 such episodes.
  - Typical episode lasts 6 quarters; housing busts usually last more than 16 quarters.
  - Typical (median) decline in house prices is 6 percent; average decline around 11 percent due to large declines in sample.
  - During a house price bust, prices decline by about 29 percent typically.
- Macroeconomic effects:
  - Output typically still expands during house price decline episodes, reflecting long duration.
  - Residential investment typically shrinks by 4 percent during declines and 12 percent during busts.
  - Total investment typically goes down by more than 8 percent.
  - Unemployment records a statistically significant increase during bust episodes relative to non-bust episodes.
  - Inflation tends to be much lower at the end of house price busts, by some 3 percentage points.
  - Credit still expands over house price decline episodes but at a slower rate than normal; equity prices do not change much.
- Dynamics around house price busts:
  - Slowdown in output around a house price bust is more gradual than in a credit crunch.
  - Investment declines occur after the onset and involve both residential and nonresidential investment.
  - Residential investment recovery takes much longer in house price busts than in credit crunches.
  - Inflation experiences a sharp decline after a few quarters; unemployment starts to rise after about two years.
  - House prices remain on the decline for much more than three years.
  - Equity prices usually begin to recover within two years; credit growth experiences a large slowdown and does not return to pre-bust levels for at least three years.

C. Episodes of Declines in Equity Prices
- Counts and durations:
  - 234 equity price decline episodes; 58 busts.
  - Typical country had around 11 declines and 3 busts.
  - Typical decline lasts 5 quarters and is associated with a price drop of 27 percent.
  - Equity busts typically last 10 quarters and are accompanied with a 50 percent price decline.
- Macroeconomic effects:
  - Output and consumption continue to grow during equity declines but at lower rates.
  - No decline in investment over equity decline episodes (aggregate finding), though later discussion shows nonresidential investment dynamics differ.
  - Unemployment picks up slightly; inflation does not change much.
  - Credit still registers expansion; house prices typically increase between the peak and trough of equity decline episodes.
- Dynamics around equity busts:
  - Output slowdown usually starts three quarters after start of the equity bust; level of output typically does not decline.
  - Slowdown in consumption delayed until after one year and weaker than in credit crunches or house price busts.
  - Investment growth slows only after 3 to 4 quarters; non-residential investment initially increases for a few quarters before falling faster than residential investment.
  - Equity price falls are sharp and prolonged; prices do not start to recover within three years following the start.
  - Credit growth experiences a delayed slowdown, picking up somewhat two years after the beginning of the equity bust.
  - House price growth rate typically starts to decline only after one year, becoming negative after two years.

D. Credit Contractions and Asset Price Declines: Summary
- Duration and amplitude comparisons:
  - House price declines and busts last longer than credit contractions/crunches or equity price declines/busts.
  - Equity price drops are less persistent but much larger in amplitude.
  - Typical episodes:
    - House price decline (bust): 6 (29) percent drop in house prices.
    - Equity price decline (bust): 27 (50) percent fall in equity prices.
  - Residential investment declines by 6 percent during credit crunches and 12 percent during house busts.
- Recessions associated with crunches/busts:
  - Recessions associated with credit crunches, house price busts, and equity price busts were identified: 18, 34 and 45 episodes respectively.
  - Severe crunch/bust episodes defined as top 12.5 percent (or top half of crunches/busts depending on context) are analyzed for distributional features.

### V. Recessions Associated with Increases in Oil Prices
- Definition: oil price "shock" (severe oil price shock) if increase is in the top quartile (12.5 percent) of all price increases.
- Finding: almost half of the recessions in the sample are associated with an oil price shock.
- Differences: recessions associated with severe oil price shocks exhibit significantly larger output drops (detailed tabular results referenced in source).

*Source: _wp08274 - 0.03 percent per quarter, but the rate of growth during expansions was close to 0.75 percent.*

### 2.6 percent versus 1.8 percent for those recessions without an oil price increase. Likewise,

### _wp08274 - 2.6 percent versus 1.8 percent for those recessions without an oil price increase. Likewise,

### Oil price shocks and stagflationary recessions
- Recessions coinciding with oil price shocks:
  - Output decline: "2.6 percent versus 1.8 percent" for those recessions without an oil price increase.
  - Consumption, residential investment and industrial production register noticeably greater declines than in recessions without oil price shocks.
  - Both imports and exports fall significantly during these recessions.
  - Equity and house price changes are not much different than in recessions without jumps in oil prices.
  - Inflation behaviour:
    - While the rate of inflation falls in a typical recession, recessions coinciding with oil price shocks are accompanied by significantly higher inflation rates.
    - These episodes are mainly stagflationary in type.
    - "Inflation was usually high, and in some cases even accelerating, during most of the recessions associated with oil price shocks."
    - Over "one third" of recessions in the sample show an increase in the rate of inflation; those in the top quartile of inflation changes are categorized as inflation "shocks".
- Stagflationary recessions (recessionary periods accompanied by a pick up in the inflation rate):
  - Similar to recessions with oil price shocks: significantly larger decline in output and, by construction, a much larger jump in the rate of inflation than other recessions.
  - Equity prices fall significantly more in recessions with an acceleration in inflation.
  - Other financial variables do not show statistically significant different changes relative to other recessions.

### Policy responses during recessions, crunches and busts (monetary and fiscal proxies)
- Scope and proxies:
  - Monetary policy proxied by changes in short-term interest rates (nominal and real).
  - Fiscal policy proxied by changes in real government consumption.
  - Recognized limits: these are narrow, coarse proxies and associations may not be causal.
- General patterns:
  - Both monetary and fiscal policies tend to be countercyclical during recessions, credit contractions and asset price declines.
  - Fiscal policy appears more accommodative in severe recessions, credit crunches and asset price busts.
- Specific findings:
  - In episodes with credit crunches, house price and equity price busts, government consumption rises significantly more than in other contraction and bust episodes—suggesting more aggressive countercyclical fiscal policy when credit frictions are present.
  - During house price busts, the decline in nominal interest rates is statistically significantly larger than in episodes without house price busts.
  - The growth rate of government consumption increases to twice that in recessions without crunches (statistically significant for recessions with credit crunches).
  - Recessions with oil price shocks and those with a jump in inflation:
    - Drops in short-term real interest rates are larger; differences are statistically significant.
    - Nominal short-term interest rate stays constant in recessions with a large increase in inflation, while it decreases in other recessions; this difference is statistically significant.
    - Interpretation: in stagflationary recessions nominal interest rate increases did not keep up with inflation increases.

### Recession outcomes and financial factors — determinants of recession costs
- Regression framework and sample:
  - Baseline OLS regressions use a large sample of recessions over the period "1960:1-2007:4".
  - Financial regressors include changes in credit, housing and equity prices during recessions.
  - Controls include cumulative output growth over the two years preceding the recession, change in exports during recessions (external demand), growth rate of oil prices in the two years preceding the recession, changes in government expenditures, changes in short-term real interest rates, a "Great Moderation" dummy (value of one after "1986:2", zero otherwise), and a crisis dummy (banking or currency crisis during or in the year prior to a recession).
  - Regressions pick up associations, not necessarily causalities.
- Key empirical findings:
  - Coefficients on financial variables are positive: declines in credit, house prices and equity prices are positively associated with the depth of recessions; house and equity prices initially statistically significant.
  - Housing prices:
    - Decline in housing prices appears more influential in determining the cost of recessions than contraction in credit or drop in equity prices.
    - In specifications including all financial variables, the coefficient on housing prices remains statistically significant and positive.
    - In several specifications the coefficient on equity prices loses significance; credit’s coefficient can change sign depending on specification.
    - Augmenting regressions with recession duration preserves the significantly positive influence of housing prices.
    - Overall: "the change in house prices tends to be the most robust financial variable associated with the depth of recessions."
  - Other determinants:
    - State of the economy at onset (cumulative growth prior two years) is positively associated with the extent of declines—stronger expansions are followed by larger contractions.
    - Decline in exports is positively correlated with recession depth and is significant in almost all specifications.
    - The "Great Moderation" dummy indicates milder recessions in most specifications.
    - Neither change in oil prices nor presence of a financial crisis appear to affect severity in baseline specifications; this may reflect that financial variable changes capture crisis effects, though results vary in some specifications.
    - Recession amplitude is positively associated with its duration.
- Robustness and additional checks:
  - Quantile regressions (Tables 14A-14B) preserve main findings: changes in housing prices remain significantly positively correlated with recession costs.
  - In one specification the crisis dummy becomes positive and statistically significant, suggesting a positive association between financial crises and recession costs.
  - Additional sensitivity tests (fixed effects, distributional checks) do not change main messages.

### Conclusions and policy-relevant lessons
- Summary of findings:
  - Interactions between macroeconomic and financial variables are key to determining severity and duration of recessions.
  - Recessions associated with credit crunches and house price busts are deeper and longer than other recessions.
  - Credit crunches and house price busts have longer durations than typical recessions and lead to sharper declines in consumption and investment; house price busts especially linked to larger consumption declines via housing wealth losses.
  - Equity price busts are less consistently associated with real sector outcomes.
  - Across specifications, changes in house prices most consistently influence recession depth.
  - Control variables consistently influencing recession depth: state of economy at onset and recession duration. Recessions during Great Moderation are milder.
- Lessons for the global financial storm referenced:
  - Recessions following the financial storm are likely to be more costly because they coincide with simultaneous credit crunches and asset price busts.
  - Global dimensions of such crises are likely to intensify in coming months (per the source).
  - The severity for any country depends on pre-recession financial health of firms, banks, households, and on policy measures taken.
  - Continued decisive policy actions at national and global levels could help meet evolving challenges.

### Caveats and agenda for future research
- Caveats:
  - Event-study nature: no causal inferences about how recessions originate, direction of causation between financial variables and macro outcomes, nor definitive statements on policy potency.
  - Initial conditions, external demand and supply developments, and policy responses influence recession paths; regressions attempt to control but cannot fully isolate effects.
  - Proxies for fiscal and monetary policies are coarse and policy effects may operate with lags.
- Future research directions proposed:
  - Explore channels through which financial and real variables interact (wealth, substitution, balance sheet effects).
  - Use individual firm data across countries to examine firm-level financial variables (credit use, inventory, liquidity), heterogeneity by firm size and leverage, and sensitivity to external finance constraints.
  - Focus on alternative metrics of activity (various output gap measures) to study patterns in recessions associated with financial stress or crises.
  - Expand sample to include emerging market economies to examine global dimensions of recessions.

*Source: IMF working paper content (excerpt).*

### REFERENCES

### _wp08274 - REFERENCES

### Major bibliographic sources
- Key academic and policy contributions cited (authors and topics as listed): Artis; Backus, Kehoe, Kydland; Bernanke; Blanchard; Borio; Bry & Boschan; Burns & Mitchell; Calvo et al.; Canova; Cardarelli et al.; Carroll et al.; Cecchetti; Cerra & Saxena; Claessens et al.; Cochrane; Cotis & Coppel; Crucini et al.; Dell’Ariccia & Garibaldi; Ferguson; Fisher; Gordon; Hall; Hamilton; Harding & Pagan; Hansen & Prescott; Helbling & Terrones; Kashyap & Stein; Kehoe & Prescott; Kilian; Kiyotaki & Moore; Kose et al.; Kydland & Prescott; Leamer; Lown & Morgan; Mendoza & Terrones; Mendoza; Mishkin; Morsink et al.; Muellbauer; Obstfeld & Rogoff; Pagan & Sossounov; Perry et al.; Plosser; Reinhart & Rogoff; Romer; Romer & Romer; Stock & Watson; Terrones; Zarnowitz.

### Dataset coverage and variables (sample periods and sources preserved)
- Output: Gross domestic product, volume; 1960:1-2007:4* — OECD.
- Consumption: Private final consumption expenditure, volume; 1960:1-2007:4 — OECD.
- Government Consumption: Government final consumption expenditure, volume; 1960:1-2007:4 (except Spain: 1961:1-2007:4) — OECD.
- Investment: Gross fixed capital formation, volume; 1960:1-2007:4 — OECD.
- Residential FCF: Private residential fixed capital formation, volume; 1960:1-2007:4 (except Canada: 1961:1-2007:4, France: 1963:1-2007:4, New Zealand: 1961:3-2007:4, Portugal: 1988:1-2007:4) — OECD.
- Nonresidential FCF: Private nonresidential fixed capital formation, volume; 1960:1-2007:4 (country exceptions listed) — OECD.
- Total FCF: Private total fixed capital formation, volume; 1960:1-2007:4 (country exceptions listed) — OECD.
- Industrial Production: Industrial production; 1960:1-2007:4 — IFS (coverage note included).
- Exports / Imports: Exports/imports of goods and services, volume; 1960:1-2007:4 — OECD.
- Export Prices / Import Prices: Export/import unit values; 1960:1-2007:4 — IFS (country exceptions listed).
- Net Export–GDP ratio: Net exports/GDP; 1960:1-2007:4 — Both net exports and GDP are from OECD (France exception noted).
- Current Account–GDP Ratio: Current account balances/GDP; 1960:1-2007:4 — (1) Current account balances are from OECD and GDS (country exceptions listed); (2) GDP is from OECD.
- NEER: Nominal effective exchange rate; 1960:1-2007:4 (country exceptions listed) — IFS.
- REER: Real effective exchange rate; 1980:1-2007:4 — IFS.
- House Prices: Nominal house prices deflated using CPI (BIS data only); 1970:1-2007:4 (country exceptions listed) — OECD and BIS (Austria, Belgium, Greece and Portugal).
- Stock Prices: Share Price (Index) deflated using Consumer Price Index; 1960:1-2007:4 (country exceptions listed) — IFS.
- Real Credit: Nominal credit deflated using Consumer Price Index; 1960:1-2007:4 (country exceptions listed) — (1) IFS and Datastream for nominal credit; (2) CPI from IFS.
- Short-term Real Interest Rate: Treasury bill rate deflated using inflation rate; 1960:1-2007:4 (except Australia 1969:3-2007:4) — (1) Short-term nominal interest rate is from IFS; (2) Inflation rate is annual CPI growth from IFS.
- Long-term Real Interest Rate: Government bond yield deflated using inflation rate; 1960:1-2007:4 (country exceptions listed) — IFS.
- Unemployment Rate: Unemployment rate; 1960:1-2007:4 — OECD, GDS, HAVER, DATASTREAM and BLOOMBERG.
- Inflation Rate: Inflation rate; 1960:1-2007:4 — CPI from IFS; inflation rate formula preserved: [CPI(quarter i, year t)/CPI(quarter i, year t-1)-1]*100, where i=1,2,3,4.
- US series note: The series for US is from 1960:1-2008:1; same for all other series.

### Key empirical tallies and event counts (preserved exactly)
- Total recessions in sample: 122.
- Recessions associated with crunches and busts: 76.
- Recessions associated with none: 46.
- Of 122 recessions:
  - 18 are associated with credit crunches.
  - 34 are associated with house price busts.
  - 45 are associated with equity price busts.

### Stylized recession and event statistics (medians / means and exact metrics preserved)
- Recessions (OECD group):
  - Median Number of Recessions: 5.00.
  - Median Duration (quarters): 3.00.
  - Median Proportion of time in recession: 0.11.
  - Median Amplitude: -1.87.
  - Median Cumulative Loss: -3.04.
  - Mean Number of Recessions: 5.81.
  - Mean Duration: 3.64.
  - Mean Proportion of time in recession: 0.11.
  - Mean Amplitude: -2.63.
  - Mean Cumulative Loss: -6.40.
- Recession median/mean comparisons for G-7 and Other country groups provided in Table 1.A (values preserved in tables).

- Aggregate event durations and amplitudes (Table 5 summary, exact cells):
  - Credit Contractions (Mean duration): 5.65; Median amplitude: -4.22.
  - Credit Crunches (Mean duration): 10.3***; Median amplitude: -17.03***.
  - House Price Declines (Mean duration): 8.47; Median amplitude: -5.99.
  - House Price Busts (Mean duration): 18.43***; Median amplitude: -28.52***.
  - Equity Price Declines (Mean duration): 6.91; Median amplitude: -26.58.
  - Equity Price Busts (Mean duration): 11.78***; Median amplitude: -50.27***.
  - Residential Investment, Non-Residential Investment, Unemployment median responses to events are preserved in Table 5.

- Leads and lags between events and recessions (Table 6):
  - Median lead (quarters between start of crunch/bust and start of recession):
    - Credit Crunches: 4.00.
    - House Price Busts: 3.00.
    - Equity Price Busts: 5.00.
  - Median lag (quarters between end of recession and end of crunch/bust):
    - Credit Crunches: 2.00.
    - House Price Busts: 9.00.
    - Equity Price Busts: 0.00.

### Comparative impacts on recessions by event type (selected median/mean metrics preserved)
- Recessions associated with Credit Crunches (Table 7):
  - Median Amplitude: -2.19 (with crunches) vs. -1.82 (without).
  - Median Cumulative Loss: -4.44 (with) vs. -2.87 (without).
  - Median Credit change: -4.25 (with crunches) vs. 1.54 (without).
  - Median House Prices: -4.04 (with) vs. -1.82 (without).
  - Median Equity Prices: -2.47 (with) vs. -6.28 (without).

- Recessions associated with House Price Busts (Table 8):
  - Median Duration: 3.20 (with busts) vs. 3.00 (without).
  - Median Amplitude: -2.18 (with) vs. -1.52 (without).
  - Median Cumulative Loss: -3.74 (with) vs. -2.25 (without).
  - Median Consumption: -0.73 (with) vs. 0.09 (without).
  - Median Credit: -0.50 (with) vs. 2.42 (without).
  - Median House Prices: -6.28 (with) vs. -0.82 (without).

- Recessions associated with Equity Price Busts (Table 9):
  - Median Amplitude: -1.98 (with) vs. -1.63 (without).
  - Median Cumulative Loss: -3.08 (with) vs. -2.64 (without).
  - Median Total Investment: -6.17 (with) vs. -3.17 (without).
  - Median Credit: 1.00 (with) vs. 1.06 (without).
  - Median Equity Prices: -13.05 (with) vs. -0.80 (without).

- Recessions and Oil Price Shocks (Table 11):
  - Median Amplitude: -2.05 (with oil shocks) vs. -1.78 (without).
  - Median Cumulative Loss: -3.14 (with) vs. -2.94 (without).
  - Median Inflation Rate change: 0.29 (with) vs. -0.90 (without) — note significance marks preserved in table.

### Policy variable changes during events (median values preserved, Table 12)
- Recessions (median changes):
  - Short-Term Nominal Interest Rate: -0.79.
  - Short-Term Real Interest Rate: -0.70.
  - Government Consumption (percent change): 1.79.
- Severe Recessions (median changes):
  - Short-Term Nominal Interest Rate: 0.00.
  - Short-Term Real Interest Rate: -1.11.
  - Government Consumption: 2.16.
- Credit Crunches (median changes):
  - Short-Term Nominal Interest Rate: -1.50.
  - Short-Term Real Interest Rate: -0.09.
  - Government Consumption: 6.33***.
- House Price Busts (median changes):
  - Short-Term Nominal Interest Rate: -3.16***.
  - Short-Term Real Interest Rate: 0.21.
  - Government Consumption: 9.07***.
- Equity Price Busts (median changes):
  - Short-Term Nominal Interest Rate: 0.28.
  - Short-Term Real Interest Rate: -0.57.
  - Government Consumption: 7.72***.
- Recessions with Credit Crunches (median government consumption): 3.84*** vs. Recessions without Credit Crunches: 1.60.

### Empirical regressions on determinants of recession amplitude (selected coefficients preserved)
- OLS regressions (Table 13.A / 13.B):
  - House Price coefficients: positive and statistically significant in multiple specifications, e.g., 0.174*** (column with robust std. errors) and 0.165*** in alternate specification.
  - Exports coefficient: positive and frequently significant, e.g., 0.109***.
  - Initial Output: positive coefficients, e.g., 0.177**, 0.191*.
  - Great Moderation: negative coefficients, e.g., -0.803, -0.885* (significance preserved per table).
  - Credit coefficients mixed; in some specifications small positive, in others negative and significant (e.g., -0.090*** in some OLS columns).
  - Duration of Recession: positive and significant in specifications where included (e.g., 0.261**; 0.297**).
  - Number of observations and adjusted R-squared values preserved per column (e.g., Number of observations: 11795, 10995, 11099 etc. as shown).

- Quantile regressions (Table 14.A / 14.B):
  - House Price coefficients remain positive and significant across quantiles, e.g., 0.113***, 0.097***.
  - Credit coefficients in quantile regressions often negative and significant in certain columns, e.g., -0.054**, -0.075***.
  - Exports and Initial Output coefficients positive and significant in many quantile specifications.
  - Great Moderation coefficients negative and significant in multiple quantile specifications (e.g., -1.177**, -1.025*).
  - Financial Crisis indicator shows positive coefficients in some quantile columns (e.g., 0.561; 0.501) with varying significance.

### Figures and visual summaries (notes preserved)
- Figure 1: Associations among 122 recession episodes and overlap with credit crunches, house price busts, equity price busts (counts preserved).
- Figures 2–8: Median and quartile profiles for duration, amplitude, and time-paths of output, consumption, investment, industrial production, trade, inflation, unemployment, house prices, equity prices, credit around recessions, credit crunches, house price busts, equity price busts. Notes emphasize that solid line = median, dotted lines = upper and lower quartiles; zero denotes the peak (quarter after which recession/crunch/bust begins) and x-axis in quarters.

*Source: _wp08274 - REFERENCES (content unit provided).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2008/_wp08274.pdf_