## 3. Amplitude, Duration, and Slope of a Downturn

## Source details

**Canonical URL:** [3. Amplitude, Duration, and Slope of a Downturn](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp1414.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp1414.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp1414.pdf.json)

---

### Introduction: question and motivation
- Examines how bank-level financial soundness indicators (FSIs) relate to financial (credit) cycles, focusing on capital-asset ratio, liquid-asset ratio, and net open position (NOP)-capital ratio.
- Key hypothesis: raising FSIs during downturns can exacerbate credit contractions because banks increase capital-asset and liquid-asset ratios by shrinking assets or shifting asset composition; conversely, stronger FSIs at the onset of a downturn may make the ensuing downturn milder.
- Empirical strategy: use monthly SRF-based FSIs (January 2002–March 2012) and identify credit cycle phases with the Bry–Boschan Quarterly (BBQ) algorithm applied to monthly private sector credit volume (with HP-filtered “credit growth cycle” defined from filtered series).

### Data and FSIs
- FSIs constructed from IMF Monetary and Financial Statistics (SRFs):
  - Capital-asset ratio = total capital / total assets (not risk-weighted).
  - Liquid-asset ratio = liquidity / total assets; liquidity includes cash, deposits, short-term Treasury bills.
  - NOP-capital ratio = deposit takers' net open position in foreign exchange / total capital.
- Coverage and sample sizes (summary as presented in Table 1):
  - Capital-asset ratio: 118 economies used for summary statistics; Table 1 uses 1830 observations for Advanced, 7149 for Emerging, 4585 for Low income.
  - Liquid-asset ratio: 1707 observations for Advanced, 7021 for Emerging, 4583 for Low income.
  - NOP-capital ratio: 1657 observations for Advanced, 6543 for Emerging, 4614 for Low income.
- Selected summary statistics (from Table 1):
  - Capital-asset ratio:
    - Advanced mean 0.065, Std dev. 0.023, Min 0.010, Max 0.160.
    - Emerging mean 0.100, Std dev. 0.057, Min -0.734, Max 0.314.
    - Low income mean 0.112, Std dev. 0.054, Min -0.089, Max 0.326.
  - Liquid-asset ratio:
    - Advanced mean 0.278, Std dev. 0.096, Min 0.061, Max 0.504.
    - Emerging mean 0.365, Std dev. 0.145, Min 0.077, Max 0.924.
    - Low income mean 0.408, Std dev. 0.147, Min 0.073, Max 0.856.
  - NOP-capital ratio:
    - Advanced mean -0.463, Std dev. 1.112, Min -3.445, Max 3.783.
    - Emerging mean 0.094, Std dev. 8.464, Min -262.740, Max 288.629.
    - Low income mean 0.548, Std dev. 1.401, Min -19.846, Max 34.057.
- Comparison with IMF-FSI quarterly series (Table 2):
  - Positive and significant correlations for capital-asset ratio and liquid-asset ratio when IMF-FSIs are extrapolated to monthly frequency.
  - Regression R2: capital-asset ratio 0.53; liquid-asset ratio 0.19; NOP-capital ratio R2 = 0.012 (insignificant).

### Methods: cycle identification and empirical approach
- Cycle identification:
  - Turning points defined on monthly log credit volume series: a turning point at t must be local maximum/minimum over at least a 10-month window and a phase must be longer than five months (Harding and Pagan (2002) Bry–Boschan approach).
  - Identify “credit growth cycle” by HP-filtering the series and applying the same turning-point rules; upturns/downturns of the growth cycle are called surges/slowdowns.
- Empirical model:
  - Panel VAR (first-order) with country fixed effects (Helmert procedure) to allow endogeneity among GROWCREDIT (year-on-year credit growth), GROWCAP, GROWLIQUID, GROWNOP and controls ∆r (year-on-year change in average real lending rate) and ∆ex (year-on-year change in exchange rate).
  - Separate estimations for subsamples in downturn phases and upturn phases to compare responses.

### Effects of FSIs on credit growth (Panel VAR results)
- Main empirical findings (Tables 3a, 3b, Figure 2):
  - During downturns (Table 3a, N = 532):
    - Lagged GROWCAP: coefficient -0.01* (standard error 0.00) — increase in capital-asset ratio associated with further negative response in credit growth.
    - Lagged GROWLIQUID: coefficient -0.10*** (standard error 0.04) — increase in liquid-asset ratio strongly associated with slower credit growth.
    - Changes in GROWNOP largely insignificant for credit growth.
    - Controls: ∆r positive and significant for GROWCREDIT 0.05*** (0.01); ∆ex negative and significant for GROWCREDIT -0.08*** (0.02).
  - During upturns (Table 3b, N = 2356):
    - Lagged GROWLIQUID: coefficient 0.03*** (standard error 0.01) — increase in liquid-asset ratio associated with faster credit growth.
    - GROWCAP effect on credit growth during upturns not significant.
    - Controls: ∆r negative and significant for GROWCREDIT -0.02*** (0.00).
- Variance decomposition (Table 4) — percent of variation in GROWCREDIT 10 periods ahead explained by shocks:
  - Downturn: GROWCREDIT 0.043, GROWCAP 0.239, GROWLIQUID 0.586, GROWNOP 0.015, (rest: 0.106, 0.011).
  - Upturn: GROWCREDIT 0.020, GROWCAP 0.102, GROWLIQUID 0.795, GROWNOP 0.001, (rest: 0.080, 0.002).
  - Interpretation: liquid-asset ratio shocks explain about 10 percent of variations in credit growth during downturns and about 8 percent during upturns (as summarized in the text); table entries show relative contributions in the VAR decomposition over 10-month horizon.

### FSIs and cycle characteristics: amplitude, duration, slope
- Definitions:
  - Amplitude = absolute vertical distance between peak and trough of a phase.
  - Duration = time between turning points (in months).
  - Slope = amplitude / duration (sharpness).
- Summary statistics (Table 5):
  - Cycle downturn: observations 87, Average duration 9 months, average amplitude 0.10, average slope 0.01.
  - Cycle upturn: observations 63, Average duration 27 months, average amplitude 0.34, average slope 0.01.
  - Growth cycle slowdown: obs 300, Average duration 17 months, amplitude 0.14, slope 0.01.
  - Growth cycle surge: obs 327, Average duration 16 months, amplitude 0.14, slope 0.01.
- Duration (Weibull duration model with country fixed effects; Table 6):
  - Positive duration dependence (ln(p) significant and positive).
  - Liquid-asset ratio:
    - Downturn: LIQUID coefficient 11.43*** (3.81) — higher liquid-asset ratio at beginning of downturn associated with shorter downturn durations.
    - Upturn: LIQUID coefficient 11.50*** (3.84) — higher liquid-asset ratio associated with longer upturns.
    - Growth cycle slowdown: LIQUID 9.30*** (1.94) — higher liquid associated with shorter slowdowns.
    - Growth cycle surge: LIQUID -2.91* (1.71) — higher liquid associated with longer surges (negative coefficient here implies differing sign convention in model).
  - NOP:
    - Upturn: NOP -0.90*** (0.35) — higher NOP-capital ratio associated with longer upturns.
  - Capital-asset ratio generally insignificant for duration of downturns.
- Amplitude (fixed-effect regressions; Table 7):
  - Downturn amplitude:
    - CAP coefficient -3.04*** (0.88) — higher capital-asset ratio at beginning associated with smaller downturn amplitudes.
    - LIQUID -0.12 (0.21) (insignificant in some specs) but for growth slowdowns LIQUID -0.65*** (0.12) — higher liquid associated with smaller slowdowns.
    - NOP -0.07*** (0.02) — higher NOP-capital ratio associated with lower amplitudes of downturns/slowdowns.
    - absNOP 0.06** (0.03) — larger absolute NOP associated with larger downturn amplitude (exposure to FX risk worsens amplitude).
  - Upturn amplitude: FSIs generally insignificant for upturn amplitudes.
- Slope (fixed-effect regressions; Table 8):
  - Downturn slope:
    - CAP -0.50** (0.22) — higher capital-asset ratio associated with smaller (less sharp) downturn slopes.
    - LIQUID -0.07 (0.05) — in some specifications higher liquid-asset ratio moderates downturn sharpness.
    - absNOP 0.02*** (0.00) — higher absolute NOP associated with faster contraction (sharper slope).
  - Upturn slope:
    - CAP sometimes associated with smaller slopes for upturns (less rapid expansion).
    - NOP -0.01** (0.00) — higher NOP-capital ratio related to smaller slopes of upturns and surges (limits speed of expansion).
- Synthesis of cycle-characteristics findings:
  - Stronger FSIs (higher capital-asset ratio, higher liquid-asset ratio, higher NOP-capital ratio and lower abs(NOP)) at onset of a downturn are associated with more moderate downturns — smaller amplitude and smaller slope — and with shorter downturn durations (mainly via liquid-asset ratio).
  - Upturns that start with higher FSIs tend to last longer and be less abrupt.

### Policy implications and conclusion
- Timing matters: strengthening banks’ FSIs is most effective as a macro-stabilizing strategy when done during booms (salient credit expansions) so banks enter downturns better capitalized and more liquid.
- Caution when tightening FSIs during downturns: empirical evidence shows that increases in capital-asset ratio and liquid-asset ratio during contraction phases are associated with further declines in credit growth (trade-off between lending and improving FSIs when liquidity is scarce and capital raising is difficult).
- Net effect: while attempting to raise FSIs in a downturn may deepen a credit contraction, having stronger FSIs in place prior to a downturn tends to produce milder and less sharp contractions and longer, less abrupt expansions.
- Recommended orientation: use FSIs proactively (countercyclically) during booms to reduce the severity and volatility of subsequent downturns; avoid policies that tighten FSIs in the midst of recessions that could unintentionally amplify credit contractions.

*Source: IMF working paper chapter "3. Amplitude, Duration, and Slope of a Downturn" (SRF-FSIs, monthly data January 2002–March 2012; tables and regression results as presented in the source PDF).*

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

### _wp1414 - References

### Tables
- 1. Summary Statistics of SRF-FSIs .........................................................................................19
- 2. Regression of SRF-FSIs on IMF-FSIs (with country fixed effects) ....................................19
- 3a. Panel VAR: Credit Growth and Changes in SRF-FSIs During Downturns.......................20
- 3b. Panel VAR: Credit Growth and Changes in SRF-FSIs During Upturns ...........................20
- 4. Variance Decomposition for Credit Growth ........................................................................21
- 5. Summary Statistics of Cycle Properties ...............................................................................21
- 6. Cycle Durations and SRF-FSIs ............................................................................................22
- 7. Cycle Amplitudes and SRF-FSIs .........................................................................................23
- 8. Cycle Slopes and SRF-FSIs .................................................................................................24

### Figures
- 1. Development of FSIs for Selected European Countries, 2005–2011 ....................................8
- 2. Impulse Responses of Credit Growth to Shocks in Growth in Capital-asset Ratio,  
    Liquid-asset Ratio, and Net Open Position-capital Ratio ....................................................13

*Source: _wp1414 - References*

### 3. Amplitude, Duration, and Slope of a Downturn ..................................................................15

### 3. Amplitude, Duration, and Slope of a Downturn

### Introduction: question and motivation
- Examines how bank-level financial soundness indicators (FSIs) relate to financial (credit) cycles, with focus on capital-asset ratio, liquid-asset ratio, and net open position (NOP)-capital ratio.
- Key hypothesis: raising FSIs during downturns can exacerbate credit contractions because banks increase capital-asset and liquid-asset ratios by shrinking assets or shifting asset composition; conversely, stronger FSIs at the onset of a downturn may make the ensuing downturn milder.
- Empirical strategy: use monthly SRF-based FSIs (January 2002–March 2012) and identify credit cycle phases with the Bry–Boschan Quarterly (BBQ) algorithm applied to monthly private sector credit volume (with HP-filtered “credit growth cycle” defined from filtered series).

### Data and FSIs
- FSIs constructed from IMF Monetary and Financial Statistics (SRFs): capital-asset ratio (total capital / total assets, not risk-weighted), liquid-asset ratio (liquidity / total assets; liquidity includes cash, deposits, short-term Treasury bills), NOP-capital ratio (deposit takers' net open position in foreign exchange / total capital).
- Coverage and sample sizes:
  - Capital-asset ratio: 118 economies used for summary statistics (Table 1 uses 1830 observations for Advanced, 7149 for Emerging, 4585 for Low income).
  - Liquid-asset ratio: 1707 observations for Advanced, 7021 for Emerging, 4583 for Low income.
  - NOP-capital ratio: 1657 observations for Advanced, 6543 for Emerging, 4614 for Low income.
- Selected summary statistics (from Table 1):
  - Capital-asset ratio: Advanced mean 0.065, Std dev. 0.023, Min 0.010, Max 0.160; Emerging mean 0.100, Std dev. 0.057, Min -0.734, Max 0.314; Low income mean 0.112, Std dev. 0.054, Min -0.089, Max 0.326.
  - Liquid-asset ratio: Advanced mean 0.278, Std dev. 0.096, Min 0.061, Max 0.504; Emerging mean 0.365, Std dev. 0.145, Min 0.077, Max 0.924; Low income mean 0.408, Std dev. 0.147, Min 0.073, Max 0.856.
  - NOP-capital ratio: Advanced mean -0.463, Std dev. 1.112, Min -3.445, Max 3.783; Emerging mean 0.094, Std dev. 8.464, Min -262.740, Max 288.629; Low income mean 0.548, Std dev. 1.401, Min -19.846, Max 34.057.
- Comparison with IMF-FSI quarterly series (Table 2): positive and significant correlations for capital-asset ratio and liquid-asset ratio when IMF-FSIs are extrapolated to monthly frequency:
  - Regression R2: capital-asset ratio 0.53; liquid-asset ratio 0.19; NOP-capital ratio R2 = 0.012 (insignificant).

### Methods: cycle identification and empirical approach
- Cycle identification:
  - Turning points defined on monthly log credit volume series: a turning point at t must be local maximum/minimum over at least a 10-month window and a phase must be longer than five months (Harding and Pagan (2002) Bry–Boschan approach).
  - Also identify “credit growth cycle” by HP-filtering the series and then applying the same turning-point rules; upturns/downturns of the growth cycle are called surges/slowdowns.
- Empirical model:
  - Panel VAR (first-order) with country fixed effects (Helmert procedure) to allow endogeneity among GROWCREDIT (year-on-year credit growth), GROWCAP, GROWLIQUID, GROWNOP and controls ∆r (year-on-year change in average real lending rate) and ∆ex (year-on-year change in exchange rate).
  - Separate estimations for subsamples in downturn phases and upturn phases to compare responses.

### Effects of FSIs on credit growth (Panel VAR results)
- Main empirical findings (Tables 3a, 3b, Figure 2):
  - During downturns (Table 3a, N = 532):
    - Lagged GROWCAP: coefficient -0.01* (standard error 0.00) — increase in capital-asset ratio associated with further negative response in credit growth.
    - Lagged GROWLIQUID: coefficient -0.10*** (standard error 0.04) — increase in liquid-asset ratio strongly associated with slower credit growth.
    - Changes in GROWNOP largely insignificant for credit growth.
    - Controls: ∆r positive and significant for GROWCREDIT (0.05*** (0.01)); ∆ex negative and significant for GROWCREDIT (-0.08*** (0.02)).
  - During upturns (Table 3b, N = 2356):
    - Lagged GROWLIQUID: coefficient 0.03*** (standard error 0.01) — increase in liquid-asset ratio associated with faster credit growth.
    - GROWCAP effect on credit growth during upturns not significant.
    - Controls: ∆r negative and significant for GROWCREDIT (-0.02*** (0.00)).
- Variance decomposition (Table 4) — percent of variation in GROWCREDIT 10 periods ahead explained by shocks:
  - Downturn: GROWCREDIT 0.043, GROWCAP 0.239, GROWLIQUID 0.586, GROWNOP 0.015, (rest: 0.106, 0.011).
  - Upturn: GROWCREDIT 0.020, GROWCAP 0.102, GROWLIQUID 0.795, GROWNOP 0.001, (rest: 0.080, 0.002).
  - Interpretation: liquid-asset ratio shocks explain about 10 percent of variations in credit growth during downturns and about 8 percent during upturns (as summarized in the text); the table entries show relative contributions in the VAR decomposition over 10-month horizon.

### FSIs and cycle characteristics: amplitude, duration, slope
- Definitions:
  - Amplitude = absolute vertical distance between peak and trough of a phase.
  - Duration = time between turning points (in months).
  - Slope = amplitude / duration (sharpness).
- Summary statistics (Table 5):
  - Cycle downturn: observations 87, Average duration 9 months, average amplitude 0.10, average slope 0.01.
  - Cycle upturn: observations 63, Average duration 27 months, average amplitude 0.34, average slope 0.01.
  - Growth cycle slowdown: obs 300, Average duration 17 months, amplitude 0.14, slope 0.01.
  - Growth cycle surge: obs 327, Average duration 16 months, amplitude 0.14, slope 0.01.
- Duration (Weibull duration model with country fixed effects; Table 6):
  - Positive duration dependence (ln(p) significant and positive).
  - Liquid-asset ratio:
    - Downturn: LIQUID coefficient 11.43*** (3.81) — higher liquid-asset ratio at beginning of downturn associated with shorter downturn durations.
    - Upturn: LIQUID coefficient 11.50*** (3.84) — higher liquid-asset ratio associated with longer upturns.
    - Growth cycle slowdown: LIQUID 9.30*** (1.94) — higher liquid associated with shorter slowdowns.
    - Growth cycle surge: LIQUID -2.91* (1.71) — higher liquid associated with longer surges (negative coefficient here implies differing sign convention in model; interpret per paper: higher liquid-asset ratio leads to longer upturns and surges).
  - NOP:
    - Upturn: NOP -0.90*** (0.35) — higher NOP-capital ratio associated with longer upturns (table shows significance; interpret in context: higher NOP-capital ratio and lower abs(NOP) milder amplitude/slope).
  - Capital-asset ratio generally insignificant for duration of downturns.
- Amplitude (fixed-effect regressions; Table 7):
  - Downturn amplitude:
    - CAP coefficient -3.04*** (0.88) — higher capital-asset ratio at beginning associated with smaller downturn amplitudes.
    - LIQUID -0.12 (0.21) (insignificant in some specs) but for growth slowdowns LIQUID -0.65*** (0.12) — higher liquid associated with smaller slowdowns.
    - NOP -0.07*** (0.02) — higher NOP-capital ratio associated with lower amplitudes of downturns/slowdowns.
    - absNOP 0.06** (0.03) — larger absolute NOP associated with larger downturn amplitude (exposure to FX risk worsens amplitude).
  - Upturn amplitude: FSIs generally insignificant for upturn amplitudes.
- Slope (fixed-effect regressions; Table 8):
  - Downturn slope:
    - CAP -0.50** (0.22) — higher capital-asset ratio associated with smaller (less sharp) downturn slopes.
    - LIQUID -0.07 (0.05) (some significance in specifications) — higher liquid-asset ratio moderates downturn sharpness.
    - absNOP 0.02*** (0.00) — higher absolute NOP associated with faster contraction (sharper slope).
  - Upturn slope:
    - CAP sometimes associated with smaller slopes for upturns (less rapid expansion): e.g., CAP 0.13 (not significant) in one spec; overall higher capital restrains speed of expansion.
    - NOP -0.01** (0.00) — higher NOP-capital ratio related to smaller slopes of upturns and surges (limits speed of expansion).
- Synthesis of cycle-characteristics findings:
  - Stronger FSIs (higher capital-asset ratio, higher liquid-asset ratio, higher NOP-capital ratio and lower abs(NOP)) at onset of a downturn are associated with more moderate downturns — smaller amplitude and smaller slope — and with shorter downturn durations (mainly via liquid-asset ratio).
  - Upturns that start with higher FSIs tend to last longer and be less abrupt.

### Policy implications and conclusion
- Timing matters: strengthening banks’ FSIs is most effective as a macro-stabilizing strategy when done during booms (salient credit expansions) so banks enter downturns better capitalized and more liquid.
- Caution when tightening FSIs during downturns: empirical evidence shows that increases in capital-asset ratio and liquid-asset ratio during contraction phases are associated with further declines in credit growth (trade-off between lending and improving FSIs when liquidity is scarce and capital raising is difficult).
- Net effect: while attempting to raise FSIs in a downturn may deepen a credit contraction, having stronger FSIs in place prior to a downturn tends to produce milder and less sharp contractions and longer, less abrupt expansions.
- Recommended orientation: use FSIs proactively (countercyclically) during booms to reduce the severity and volatility of subsequent downturns; avoid policies that tighten FSIs in the midst of recessions that could unintentionally amplify credit contractions.

*Source: IMF working paper chapter "3. Amplitude, Duration, and Slope of a Downturn" (SRF-FSIs, monthly data January 2002–March 2012; tables and regression results as presented in the source PDF).*

### REFERENCES

### REFERENCES

### References list

- Abiad, A., G. Dell’Ariccia, and B. Li, 2011, “Creditless Recoveries,” IMF Working Paper, WP/11/58.  
- Babihuga, R., 2007, “Macroeconomic and financial soundness indicators: An empirical investigation”, IMF Working Paper, WP/07/115.  
- Bikker, J.A. and P.A.J. Metzemakers, 2002, “Bank provisioning behavior and procyclicality,” Bank of Netherlands Research Series Supervision, No 50.  
- Cihak, M. and K. Schaeck, 2010, “How well do aggregate prudential ratios identify banking system problems?,” Journal of Financial Stability, 6, 130-44.  
- Claessens, S., M. A. Kose, and M. E. Terrones, 2011, “How do business and financial cycles interact?”, IMF Working Paper, WP/11/88.  
- Covas, F. and S. Fujita, 2009, “Time-varying capital requirements in a general equilibrium model of liquidity dependence,” Working Paper, Federal Reserve Bank of Philadelphia.  
- Forbes, K. and F. E. Warnock, 2011, “Capital flow waves: Surges, stops, flight and retrenchment,” Working Paper.  
- Harding, D. and A. Pagan, 2002, “Dissecting the cycle: A methodological investigation”, Journal of Monetary Economics, Vol. 49, 365–81.  
- Harding, D. and A. Pagan, 2011, “An econometric analysis of some models for constructed binary time series,” Journal of Business and Economic Statistics, Vol. 29, No 1, 86–95.  
- Hardy, D. and C. Pazarbasioglu, 1998, “Leading indicators of banking crises: Was Asia different?,” IMF Working Paper, WP/98/91.  
- IMF (2006), Financial Soundness Indicators Compilation Guide.  
- IMF (2000), Monetary and Financial Statistics Manual.  
- IMF (2008), Monetary and Financial Statistics Manual—Compilation Guide.  
- Kaminsky, G. S. Lizondo, and C. M. Reinhart, 1998, “Leading indicators of currency crises,” IMF Staff Papers, Vol. 45, No 1.  
- King, R. G., and C. I. Plosser, 1994, “Real business cycles and the test of the Adelmans,” Journal of Monetary Economics, Vol. 33, No 2, 405–38.  
- Love, I. and L. Zicchino, 2006, “Financial development and dynamic investment behavior: Evidence from panel VAR,” The Quarterly Review of Economics and Finance, 46, 190–210.  
- N’Diaye, P., 2009, “Countercyclical macro prudential policies in a supporting role to monetary policy,” IMF Working Paper, WP/09/257.  
- Quagliariello, M., 2004, “Banks’ performance over the business cycle: A panel analysis on Italian Intermediaries,” University of York Discussion Paper.  
- Resende, C., A. Dib, R. Lalonde, and N. Perevalov, 2011, “Countercyclical bank capital requirement and optimized monetary policy rules,” Bank of Canada Working Paper.  
- San Jose, Krueger, and Khay (2009), “The IMF’s work on financial soundness indicators,” BIS Irving Fisher Committee Bulletin No 28.

*Source: _wp1414 - REFERENCES*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp1414.pdf_