## _wp1243

## Source details

**Canonical URL:** [_wp1243](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1243.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1243.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1243.pdf.json)

---

### Research objective and scope
- Seeks to uncover the main drivers of real credit growth in emerging Asia using a structural empirical model.
- Sample period: 1989:Q1–2010:Q4.
- Country blocks:
  - EMAS (emerging Asia): China, India, Hong Kong SAR, Korea, Singapore, Taiwan Province of China, Indonesia, Malaysia, Thailand, and the Philippines (EMAS aggregates constructed using PPP-GDP weights).
  - USEUR (foreign block): PPP-weighted average of the United States and the euro area.
- Quarterly SVAR variables: real GDP growth (log growth rates), CPI (log growth rates), real credit (log growth rates; real credit = stock of credit scaled by CPI), and level of short-term interest rates.
- Credit series source: IMF’s IFS database (IFS line 22d and line 42d, when available). Other series from IMF’s WEO and IFS databases and Haver Analytics.

### Methodology and identification strategy
- Two-block SVAR with:
  - Within-block identification: sign restrictions (monetary policy, aggregate demand, and aggregate supply shocks identified within each block underpinned by a three-equation New Keynesian framework).
  - Across-block identification: recursive (block-Cholesky) structure, with shocks in the USEUR block able to affect EMAS contemporaneously, while shocks from EMAS affect USEUR with a one-period lag (baseline specification).
- Computational strategy:
  - Joint use of sign restrictions within blocks and block-Cholesky across blocks facilitates practical use of the Fry and Pagan (2010) median target (MT) method to resolve the multiple models problem.
  - Computational runtime advantage: "about an hour or two, rather than a day or two".
- Practical implementation details:
  - Uses Fry-Pagan MT method to report impulse responses with 90 percent bands, the median, and the MT median; variance decompositions when identified by sign restrictions are presented using the MT method only.
  - Lag length: "turns out to be one."
  - System dimensions: "seven variables in a two-country system"; B referenced as a (7×7) matrix; ܿ is a (7×1) vector; ܣ_t is a (7×7) matrix; ߝ_t is a (7×1) vector.
  - Initially identifies six structural shocks (monetary policy, aggregate supply, aggregate demand for each block) following Uhlig (2005); recursive structure used for robustness to identify all seven shocks.

### Key empirical findings (variance decompositions, impulse responses)
- Dominance of domestic factors for EMAS real credit growth:
  - Domestic monetary, aggregate demand and supply shocks explain 53 percent of the variation in EMAS real credit growth.
  - External counterparts of these shocks account for 16 percent of the variation.
  - Domestic aggregate demand shocks account for 37 percent of EMAS real credit variability.
- Role of monetary policy shocks:
  - Domestic monetary policy shocks explain 7 percent of the total variation in EMAS real credit growth.
  - External monetary policy shocks explain 4 percent of the total variation.
- Dynamic responses (selected observations):
  - Domestic aggregate demand (AD) shock: affects EMAS real GDP growth on impact; hump-shaped inflation reaction; domestic credit: sharp contraction lasting about two quarters.
  - Domestic monetary policy shock: credit declines rather sharply in response to a contractionary monetary policy shock over at least two quarters (Fry-Pagan MT median).
  - Foreign monetary policy shock: does not have a sizable bearing on credit but affects EMAS real GDP growth on impact and may induce adverse effects via "sudden stop" of capital flows.
- Channels through which domestic monetary policy affects credit growth:
  - Demand channel: higher interest rates suppress consumption and investment, reducing credit demand.
  - Exchange rate channel (with exchange rate flexibility): higher interest rates can appreciate the exchange rate, tighten conditions, and reduce demand for credit.
  - Balance-sheet channel: higher interest rates depress asset prices, lowering collateral values and affecting financial intermediaries’ equity, thereby curtailing lending.
- Heterogeneity and remaining variables:
  - For five of the six SVAR variables, domestic shocks account for most variation; exception: EMAS policy rates, where external factors are more influential (e.g., Hong Kong SAR peg).
  - Countries with more flexible exchange rate regimes tend to have a lower share of external factors driving credit growth; measured correlation is –0.6.
  - Subsample: role of domestic nonmonetary shocks increased in the post-2000 subsample, likely reflecting increased regional integration.
- Robustness:
  - Findings robust to alternative identification (Peersman-style sign restrictions and classical Cholesky ordering), removing sign restrictions on real credit growth, switching orderings, and increasing the horizon of sign restrictions (tested up to ݇ = 0,ڮ,3).
  - Peersman-style alternative was computationally expensive — acceptance was about seven valid draws out of over 10 million candidate draws.

### Counterfactual and scenario analyses
- Historical Scenario (2008:Q4–2010:Q4):
  - Sample average real credit growth over entire sample: 8.9 percent.
  - Growth recession threshold in figure shading: 4.6 percent.
  - Recovery averaged 7.5 percent starting end-2009.
  - Over 2008:Q4–2010:Q4 monetary policy shocks were all expansionary; these shocks corresponded to a decrease in short-term rates of about one percentage point.
  - Counterfactual: setting expansionary monetary policy shocks to zero yields average credit closer to its historical average; conclusion: monetary policy in emerging Asia has a significant impact on real credit growth.
- Forward-Looking Scenario (two-year forecast to end-2012):
  - Historical average growth threshold: 4.6 percent (growth recession marker).
  - Illustrative warning threshold for excessive real credit growth (based on Elekdag and Wu (2011)) shown as a horizontal dashed line.
  - First scenario: monetary stance maintained at 2008:Q4–2010:Q4 levels (short-term rates about 1 percentage point lower than historical average) — implies credit growth likely higher than baseline; non-symmetric 90 percent Bayesian confidence bands indicate by end-2012 there is a one in three chance that real credit growth will exceed the warning threshold.
  - Second scenario: overlays looser foreign monetary policy (2009 experience) — increases credit growth further but impact is more modest relative to domestic stance.

### Policy implications and recommendations
- Primary focus: domestic monetary policy is a central tool for influencing credit growth in emerging Asia.
- Exchange rate policy: increased exchange rate flexibility can act as a stabilizing factor by diminishing external influences on domestic credit dynamics.
- Near-term guidance: given exceptionally uncertain global growth prospects at the time of writing, a pause in monetary tightening may be appropriate for some economies, but policymakers need to remain mindful of the possibility of lingering financial imbalances over the medium term.
- Country-specific stance: policy choices should account for country-specific circumstances and balance immediate macroeconomic responses with measures to manage medium-term financial stability risks.

*Source: Executive Summary and selected sections of _wp1243 (IMF working paper).*

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

### _wp1243 - Executive Summary

### Research objective and scope
- Seeks to uncover the main drivers of real credit growth in emerging Asia using a structural empirical model.
- Sample period: 1989:Q1–2010:Q4.
- Country blocks:
  - EMAS (emerging Asia): China, India, Hong Kong SAR, Korea, Singapore, Taiwan Province of China, Indonesia, Malaysia, Thailand, and the Philippines (EMAS aggregates constructed using PPP-GDP weights).
  - USEUR (foreign block): PPP-weighted average of the United States and the euro area.
- Quarterly variables used in the SVAR: real GDP growth (log growth rates), CPI (log growth rates), real credit (log growth rates; real credit = stock of credit scaled by CPI), and level of short-term interest rates.
- Credit series source: IMF’s IFS database (IFS line 22d and line 42d, when available). Other series from IMF’s WEO and IFS databases and Haver Analytics.

### Methodological contribution and identification strategy
- Presents a two-block structural vector autoregressive (SVAR) model with:
  - Within-block identification: sign restrictions (monetary policy, aggregate demand, and aggregate supply shocks identified within each block underpinned by a three-equation New Keynesian framework).
  - Across-block identification: recursive (block-Cholesky) structure, with shocks in the USEUR block able to affect EMAS contemporaneously, while shocks from EMAS affect USEUR with a one-period lag (baseline specification).
- Rationale for two-block setup: EMAS should not be considered a small-open economy because it contains populous countries (China, India, Indonesia), has made large contributions to global growth (notably during 2009–10), affects global commodity price fluctuations, and includes global financial centers.
- Uses the Fry and Pagan (2010) median target (MT) method to resolve the multiple models problem associated with sign restrictions:
  - Reports impulse responses with 90 percent bands, the median, and the MT median.
  - For variance decompositions when identification is by sign restrictions, variance decompositions are presented using the MT method only.
- Practical advantage emphasized: the joint use of sign restrictions within blocks and a block-Cholesky recursive structure across blocks is computationally less expensive than alternatives using sign restrictions across both blocks, facilitating implementation of the Fry-Pagan MT method.

### Key empirical questions addressed
- Are domestic or external factors more important in driving emerging Asian credit growth? How important are domestic versus external monetary policies?
- Can domestic monetary policy be used to manage episodes of rapid credit growth? What other policies could help foster financial stability?
- What are the medium-term financial stability risks associated with a domestic monetary policy stance that is too loose? What are the near-term monetary policy challenges given exceptionally uncertain global growth prospects (relevant at the time of writing)?

### Main findings and quantitative conclusions
- Domestic factors are more dominant than external factors in driving rapid credit growth in emerging Asia.
  - Domestic monetary policy plays a pivotal role in guiding credit growth.
- Greater exchange rate flexibility could promote financial stability by reducing the role of external factors affecting domestic credit dynamics.
- Given the exceptionally uncertain global growth prospects at the time of writing, a pause in monetary tightening may be appropriate for some economies; however, policymakers need to remain mindful of the possibility of lingering financial imbalances over the medium term.

### Model specifics and assumptions
- Within-block shocks identified via sign restrictions follow the conceptual mapping:
  - Aggregate demand shock: positive comovement of output and prices.
  - Aggregate supply shock: negative comovement of output and prices.
  - Monetary policy shock: distinguished via the reaction of short-term interest rates (rates increased to counteract macroeconomic overheating).
- EMAS real credit growth is assumed to be procyclical in the baseline specification.
- Across-block recursive ordering in baseline: USEUR → EMAS contemporaneous effect; EMAS → USEUR with one-period lag.
- Robustness checks and sensitivity analysis reported indicate main conclusions do not depend on the baseline ordering.

### Practical and computational considerations
- The proposed joint sign-restriction + block-Cholesky identification is less computationally intensive than fully sign-restriction two-country approaches.
- The lighter computational burden enables practical use of the Fry-Pagan MT method to obtain a single set of orthogonal shocks close to medians, addressing the multiple models problem.
- Alternative identification strategies (e.g., sign restrictions only across both blocks) were found to be computationally very expensive in practice (numerous attempts over days yielded very low acceptance rates in the authors’ experiments).

### Policy implications
- Primary policy focus: domestic monetary policy is a central tool for influencing credit growth in emerging Asia.
- Exchange rate policy: increased exchange rate flexibility can act as a stabilizing factor by diminishing external influences on domestic credit dynamics.
- Near-term operational guidance: in an environment of exceptionally uncertain global growth prospects a tactical pause in monetary tightening could be warranted for some economies, but with vigilance toward medium-term financial stability risks arising from prolonged loose domestic monetary conditions.

*Source: Executive Summary of _wp1243 (IMF working paper).*

### 0.0001 percent). This is striking because the model under development is quite parsimonious

### Baseline Empirical Specification and Block-Cholesky-Sign Restrictions Identification Strategy

### Computational motivation and comparative context
- The model under development is "quite parsimonious with seven variables in a two-country system" and sign restrictions are "imposed only in the first period in the baseline specification."
- Even small-to-medium-scale VAR models identified using sign restrictions can become "computationally overbearing."
- The proposed identification scheme yields runs "in about an hour or two, rather than a day or two," facilitating use of the Fry-Pagan MT method for "meaningful variance decompositions and impulse response functions."
- Example: Peersman (2011) develops a two-country SVAR with seven variables and seven structural shocks identified using "45 sign restrictions." The text conjectures that "there were very few valid draws when all the sign restrictions were imposed jointly," motivating shock-by-shock identification in that paper and undermining simultaneous identification required for Fry and Pagan MT method.
- Consequence: Without simultaneous identification, structural shocks used for impulse responses and variance decompositions "would likely not be uncorrelated," making quantitative findings potentially not economically meaningful.

### Three other noteworthy points
- First:
  - Bjornland and Halvorsen (2008) also combine sign restrictions and short-term (zero) restrictions.
  - Differentiators of this paper:
    - (i) use of a two-country setup that is a natural case for combining sign restrictions with a recursive structure;
    - (ii) recursive ordering is across blocks, termed "block-Cholesky";
    - (iii) use of the Fry-Pagan MT method to underpin quantitative results;
    - (iv) emphasis on computational savings from combining block-Cholesky with popular sign restrictions to facilitate MT method.
- Second:
  - As in Uhlig (2005) and Helbling and others (2011), the paper initially identifies "less shocks than there are equations in the systems."
  - A seven-equation system is estimated, but "only six shocks are initially identified."
  - Rationale: In the parsimonious specification it is "not clear how such a shock would be linked to a domestic or external source," and the paper aims to uncover whether domestic or external factors drive emerging Asian credit growth.
  - Exchange rate dynamics are "partially captured by the interest rate differential between domestic and international short-term rates (as discussed in Clarida and Gali, 1994)."
  - Robustness: Using a recursive structure (which "by definition identifies all seven shocks") "reinforces our main results."
- Third:
  - Prior two-country VAR approaches (Clarida and Gali, 1994; Farrant and Peersman, 2006; Peersman, 2011) often use relative variables or symmetric/asymmetric shocks.
  - Shortcomings noted:
    - Using relative variables "implies the same propagation mechanism for shocks originating in both blocks" and "does not provide any information about the relevance of shocks for the level variables" (e.g., credit).
    - A common shock (e.g., global commodity prices) "may have opposite effects depending on whether a country is a net commodity importer or exporter," complicating interpretation in emerging Asia (where "Indonesia and Malaysia" are examples of important net exporters).
    - Identified shocks in these studies are "likely not orthogonal to each other (multiple models problem)," distorting quantitative findings.
    - Under the proposed sign restrictions, such models "may even be too large (even with seven variables) to resolve the multiple models problem using the Fry-Pagan MT method."

### C. Baseline empirical specification — model composition and notation
- Baseline two-country SVAR contains:
  - quarterly logarithmic growth rates of real GDP (both blocks),
  - CPI inflation (both blocks),
  - level of short-term interest rates (both blocks),
  - only the quarterly logarithmic growth rate of real credit for emerging Asia.
- The structural form is written as (equation (1) in source):
  - x_t = c + Σ_{i=1}^p A_i x_{t-i} + B ε_t
  - where B is a (7×7) matrix discussed below.
- Notation and dimensions preserved verbatim from source:
  - ܿ is a (7×1) vector of constants,
  - ܣ_t is a (7×7) matrix of autoregressive coefficients,
  - ߝ_t is a (7×1) vector of structural disturbances which are uncorrelated and normalized to have unit variances.
- The endogenous variables x_t included in the VAR are listed in the source (with original symbols preserved).
- Lag length ݌ is determined by "standard likelihood ratio tests and the Akaike information criterion" and "turns out to be one."

### D. Technical details — identification of B and block-Cholesky recursive structure
- Identification problem: matrix B is unidentified without further restrictions.
- Initial identification follows Uhlig (2005): identify only six structural shocks — "monetary policy, aggregate supply, and aggregate demand shocks for each of the two blocks (USEUR and EMAS)."
- Let ߗ be the estimated variance-covariance matrix of reduced-form residuals, and ܨ be the lower triangular in the Cholesky decomposition for ߗ, i.e., ܨܨ' = ߗ.
- Since ߝ_t is uncorrelated with unit variances, ߗ = B B'.
- Any two decompositions of ߗ are related by an orthonormal matrix Q; therefore ܨ̃ = B Q for some orthonormal Q.
- An arbitrary orthonormal matrix ܳ_sh can be used to construct new orthonormal matrices and new set of shocks ̃ߝ_t = ܳ_sh ε_t that are uncorrelated and have unit variances; impulse responses change with ܳ_sh.
- Block-Cholesky recursive structure:
  - The paper proposes a "block-Cholesky-sign restrictions" method for a general two-country Structural VAR(p), with block partitioning between the more "exogenous" country (X) and the other country (Z).
  - Notation preserved: blocks X_t (dimension n×1) and Z_t (dimension m×1); structural shocks ߝ_{t,1} and ߝ_{t,2} have dimensions n×1 and m×1 respectively.
  - The block representation and transformation to a normalized form are given in equations (3) and (4) in the source, with intermediate matrices ܤ_{ij} and ܪ_{ij} defined exactly as in the text.
  - For notational simplicity, ܪ_{ij} denotes ܣ_{ij}(0) with indices and dimensions preserved as in the source.

### Key assumptions (verbatim)
- Assumption 1:
  - The structural shocks (ߝ_{t,1}, ߝ'_{t,2})' ~ i.i.d. N(0, I). "This first assumption is standard in the SVAR literature, and normalizes the uncorrelated structural shocks so that they all have unit variances."
- Assumption 2:
  - "The structural coefficient matrices ܪ_{11} and ܪ_{22} are invertible."

### Explicit numeric and dimensional details (preserved exactly)
- "seven variables in a two-country system"
- Computational runtime comparison: "about an hour or two, rather than a day or two"
- Peersman (2011): "seven variables and seven structural shocks identified using 45 sign restrictions"
- Seven-equation system estimated but "only six shocks are initially identified"
- Matrix dimensions cited:
  - ܿ is a (7×1) vector
  - ܣ_t is a (7×7) matrix
  - ߝ_t is a (7×1) vector
  - B is referenced as a (7×7) matrix
- Lag length: "turns out to be one."

*Source: Excerpt from the IMF PDF chapter describing the baseline two-country SVAR, sign-restriction strategy, and block-Cholesky identification approach.*

### 3.      The structural coefficient matrix ܪ

### 3.      The structural coefficient matrix ܪ

### Identification strategy and structural setup
- Assumptions: block-Cholesky decomposition with ܪ
ଵଶ = 0; the first two assumptions imply the entire structural coefficient matrix is invertible and there is no contemporaneous effect of ߝ
௧
௒ on ܺ
௧. These assumptions, together with sign restrictions in country ܺ, identify corresponding structural shocks across the two countries.
- The block structure yields the matrix B (as in equation (1)) by partitioned-matrix inversion and leads to a reduced-form VAR representation with reduced-form residuals ݁ (equations (6)–(7)).
- Covariance structure: estimated reduced-form residual covariance matrix ߑ is partitioned according to the two blocks (external and emerging Asia). Relations (8)–(10) link elements of ܪ and ߑ, and Cholesky decompositions of the partitions produce lower-triangular factors ܨ
௑ and ܨ
௒ with orthonormal rotations ܳ
௑ and ܳ
௒ (equation (12)).

### Impulse response identification and computation
- Impulse responses of vector ݖ
௧ to a 1 standard deviation structural shock are denoted ܨܴܫ
... across horizons ݅ = 0,1,2,ڮ, and are computed via the Wold decomposition using estimated reduced-form VAR coefficients (equation (11)).
- Block-Cholesky plus third assumption implies cross-block immediate responses ܨܴܫ
௑
௘,௒(0) = 0.
- Relationships between structural shocks and residuals (7) allow rewriting impulse responses in terms of ܪ partitions, ߑ, and rotation matrices; final impulse expressions incorporate the orthonormal rotations ܳ
௑ and ܳ
௒ (equation (13)).
- Identification uses sign restrictions imposed only on impulse responses of variables to domestic shocks; across-country impulse responses are left agnostic. The procedure keeps only rotations (ܳ
௑, ܳ
௒) that produce impulse responses satisfying the sign restrictions.

### Estimation approach and algorithmic details
- Bayesian estimation: Normal-Wishart prior/posterior for the reduced-form VAR (following Uhlig (2005), Farrant and Peersman (2006), Peersman (2011)).
- For each draw from the VAR posterior:
  - Draw Householder transformations to generate rotation matrices for ܪ
ଵଵ
ିଵ and ܪ
ଶଶ
ିଵ per (12).
  - Construct impulse responses per (13).
  - Accept the draw if impulse responses satisfy all imposed sign restrictions; otherwise reject.
- Post-processing:
  - Find median impulse responses for each shock and variable.
  - Apply Fry-Pagan median target (MT) method to search for a single rotation that generates impulse responses as close to the medians as possible.

### Main empirical findings — variance decompositions and interpretive results
- Dominance of domestic factors for EMAS real credit growth:
  - Domestic monetary, aggregate demand and supply shocks explain 53 percent of the variation in EMAS real credit growth.
  - External counterparts of these shocks account for 16 percent of the variation.
  - Domestic aggregate demand shocks account for 37 percent of EMAS real credit variability.
- Role of monetary policy shocks:
  - Domestic monetary policy shocks explain 7 percent of the total variation in EMAS real credit growth.
  - External monetary policy shocks explain 4 percent of the total variation.
  - Interpretation: domestic monetary policy plays a larger role than foreign monetary policy (a proxy for global liquidity) in driving EMAS credit variation, despite external monetary policy being a channel for global liquidity.
- Channels through which domestic monetary policy affects credit growth:
  - Demand channel: higher interest rates suppress consumption and investment, reducing credit demand.
  - Exchange rate channel (with exchange rate flexibility): higher interest rates can appreciate the exchange rate, tighten conditions, and reduce demand for credit.
  - Balance-sheet channel: higher interest rates depress asset prices, lowering collateral values and affecting financial intermediaries’ equity, thereby curtailing lending.

### Remaining macro variables and heterogeneity
- For five of the six SVAR variables, domestic shocks account for most variation; the exception is EMAS policy rates, where external factors are more influential.
  - Explanation: some regional economies’ policy rates track the United States (e.g., Hong Kong SAR peg), making it difficult for the SVAR to disentangle domestic vs external monetary shocks.
- Robustness to identification:
  - First alternative (Peersman-style sign restrictions): domestic factors remain dominant. Computationally expensive — acceptance was about seven valid draws out of over 10 million candidate draws.
  - Second alternative (classical Cholesky recursive ordering): domestic credit shocks explain a large proportion of EMAS real credit growth; main conclusion that domestic factors dominate external ones remains.
  - Results are robust to removing sign restrictions on real credit growth, switching ordering in pure Cholesky decompositions, and increasing the horizon of sign restrictions (tested up to ݇ = 0,ڮ,3).
- Subsample and cross-country heterogeneity:
  - The role of domestic nonmonetary shocks (aggregate demand and supply) increased in the post-2000 subsample, likely reflecting increased regional integration.
  - Country-level variance decompositions (country-specific estimations) show domestic shocks are generally more important than external ones across economies in the region.
  - Exchange rate flexibility:
    - Countries with more flexible exchange rate regimes tend to have a lower share of external factors driving credit growth; measured correlation is –0.6.
    - Interpretation: greater exchange rate flexibility acts as a shock absorber, smoothing cyclical fluctuations affecting credit dynamics and mitigating financial imbalances.

### Policy implications and summary
- Three main policy implications drawn from variance decompositions:
  - Domestic factors are more dominant than external factors in driving rapid credit growth in Asia.
  - Domestic monetary policy accounts for a larger share of real credit variability in emerging Asia than foreign monetary policy (the external proxy for global liquidity).
  - Greater exchange rate flexibility could promote financial stability by reducing the role of external factors affecting domestic credit dynamics.
- Robustness: these findings hold across alternative identification strategies, bootstrap/rotation procedures, variable orderings in recursive schemes, and extended sign-restriction horizons.

*Source: 3. The structural coefficient matrix ܪ — _wp1243 (PDF chapter/section).*

### conclusions. Also, sign restrictions were implemented using longer time periods (in contrast

### _wp1243 - conclusions

### Impulse Response Functions
- Baseline specification: shocks identified using a combination of sign restrictions and a recursive structure across the two blocks.
- Impulse response functions shown with the 5th, 50th (the median), and 95th percentiles.
- Responses report the percentage point deviation from the mean owing to a 1 standard deviation shock.
- Medians reported using the Fry and Pagan (2010) MT method to address the multiple models problem; percentiles convey the distribution across models (not sampling uncertainty).
- Focused shocks discussed: domestic aggregate demand, domestic monetary policy, and foreign monetary policy.
- Key dynamic observations:
  - Domestic aggregate demand (AD) shock:
    - Affects EMAS real GDP growth on impact by construction (owing to imposed sign restrictions); identification scheme does not alter strength of impact.
    - Protracted effect on domestic short-term interest rates.
    - Hump-shaped inflation reaction.
    - Domestic credit: sharp contraction lasting about two quarters.
  - Domestic monetary policy shock:
    - Appears to affect credit on impact.
    - Credit seems to decline rather sharply in response to a contractionary monetary policy shock over at least two quarters as indicated by the Fry-Pagan MT median.
    - Suggests EMAS monetary policy could play a key role in managing credit growth.
  - Foreign monetary policy shock:
    - Does not have a sizable bearing on credit.
    - Seems to affect EMAS real GDP growth on impact.
    - Higher international rates may have an adverse impact on EMAS growth owing to a “sudden stop” of capital flows.
    - External funding drying up need not imply a credit drought because capital inflows may fund other asset classes (real estate, equity, corporate bonds).
    - Empirical example: credit growth increased in China during the global financial crisis, which intensified after the Lehman Brothers bankruptcy (owing to Chinese countercyclical policies).

### Counterfactual Scenarios
- Two illustrative scenarios constructed with the SVAR: a Historical Scenario and a Forward-Looking Scenario.
- Purpose: underscore the pivotal role of monetary policy in influencing credit growth in emerging Asia.

- Historical Scenario
  - Question: what if emerging Asia’s monetary policy was not expansionary once the recovery after the global financial crisis gained traction?
  - Figure 15 displays:
    - Actual evolution of emerging Asia real credit growth (bold black line).
    - Average real credit growth over the entire sample (dotted line): 8.9 percent.
    - Counterfactual path of real credit (dashed line).
    - Shaded area spanning the year starting in 2008:Q4 corresponds to the most severe phase of the global financial crisis and overlaps with growth rates lower than 4.6 percent (growth recession).
    - Starting in end-2009, the recovery gains traction and emerging Asia expands by 7.5 percent on average.
  - Monetary policy role:
    - Over 2008:Q4–2010:Q4 monetary policy shocks were all expansionary.
    - During 2008:Q4–2010:Q4 these shocks corresponded to a decrease in short-term rates of about one percentage point.
    - Counterfactual constructed by setting the expansionary monetary policy shocks to zero.
    - Result: without the monetary stimulus during the expansion, average credit would have been closer to its historical average.
    - Conclusion: monetary policy in emerging Asia has a significant impact on real credit growth.

- Forward-Looking Scenario
  - Question: how might credit growth evolve if the current emerging Asia monetary policy stance remains on hold and what are the risks?
  - Figure 16 displays:
    - Two-year forecast horizon up to end-2012 (light blue forecast).
    - Gray shaded area corresponds to a growth recession (growth below sample average of 4.6 percent).
    - Actual real credit growth (bold black line) and its historical average (dotted line).
    - Horizontal dashed line: level of real credit growth which, based on Elekdag and Wu (2011), may serve as an illustrative warning threshold above which real credit growth may be excessive.
    - Baseline model-based forecast shown with dashed black line, gradually converging to the historical average (slower after 2011).
    - First illustrative (blue) scenario: monetary stance maintained at the recovery level (2008:Q4–2010:Q4) over the two-year forecast ending in 2012; implies short-term interest rates about 1 percentage point lower than the historical average; 90 percent non-symmetric Bayesian confidence bands shown.
    - Second illustrative (orange) scenario: overlays looser foreign monetary policy (corresponding to foreign block stance during immediate aftermath of the Lehman Brothers bankruptcy).
  - Three points underscored:
    - First: barring major global economic disruptions, if the current stance of monetary policy continues, credit growth in emerging Asia is likely to be higher than the baseline and follows an upward trajectory (blue line).
    - Second: credit is more likely to grow faster, rather than slower, under this scenario as indicated by the non-symmetric 90 percent Bayesian confidence bands; by end-2012, there is a one in three chance that real credit growth in emerging Asia will exceed the warning threshold.
    - Third: an increase in global liquidity in line with the 2009 experience could exacerbate credit growth throughout the region (orange line), but the impact seems more modest.
  - Policy implication from scenarios: the monetary response to immediate macroeconomic downside risks should be balanced by measures to manage risks associated with lingering financial imbalances over the medium term; country-specific circumstances need recognition.

### Main Conclusions and Policy Implications
- Research contribution:
  - Proposes a novel identification strategy in multi-country SVARs: a two-country SVAR where shocks within blocks identified by sign restrictions and shocks across blocks by a recursive (block-Cholesky) structure.
  - Within each block (emerging Asia; United States and euro area) three shocks identified using sign restrictions: monetary policy (MP), aggregate demand (AD), and aggregate supply (AS), underpinned by a three-equation New Keynesian framework.
  - Block-Cholesky recursion allows contemporaneous influence from one block to the other, with feedback in the opposite direction occurring with a one-period lag.
  - Method is computationally less expensive than alternative identification schemes using sign restrictions, facilitating use of the Fry-Pagan MT method to resolve the multiple models problem and generate forecast error variance decompositions, impulse response functions, and other quantitative results.
- Main empirical conclusions:
  - Domestic factors are more dominant than external factors in driving rapid credit growth in Asia.
  - Domestic monetary policy plays a pivotal role in guiding credit growth.
  - Greater exchange rate flexibility could promote financial stability by reducing the role of external factors affecting domestic credit dynamics.
- Policy recommendations:
  - A pause in monetary tightening may be appropriate for some economies in view of exceptionally uncertain global growth prospects (relevant at the time of writing).
  - Policymakers should keep in mind the possibility of lingering financial imbalances over the medium term and balance monetary responses to immediate macroeconomic risks with measures to manage medium-term financial stability risks.
  - Country-specific circumstances should inform policy choices.

*Source: Authors’ calculations and analysis contained in the document.*

### REFERENCES

### REFERENCES

### Credit booms and credit shocks
- Barajas, A., G. Dell’Ariccia, and A. Levchenko, forthcoming, “Credit Booms: The Good, the Bad, and the Ugly,” IMF Working Paper (Washington: International Monetary Fund).
- Elekdag, S., and Y. Wu, 2011, “Rapid Credit Growth: Boon or Boom-Bust?” IMF Working Paper 11/241 (Washington: International Monetary Fund).
- Gourinchas, P.-O., R. Valdes, and O. Landerretche, 2001, “Lending Booms: Latin America and the World,” Economia, Vol. 1 (spring), pp. 47–99.
- Helbling, T., R. Huidrom, M.A. Kose, and C. Otrok, 2010, “Do Credit Shocks Matter? A Global Perspective,” IMF Working Paper No. 10/261 (Washington: International Monetary Fund).
- Meeks, R., 2009, “Credit Market Shocks: Evidence from Corporate Spreads and Defaults,” Working Paper 0906 (Dallas: Federal Reserve Bank of Dallas).
- Mendoza, E. and M. Terrones, 2008, “An Anatomy of Credit Booms: Evidence from Macro Aggregates and Micro Data,” NBER Working Paper No. 14049 (Cambridge, Massachusetts: National Bureau of Economic Research).
- Gilchrist, S., V. Yankov, and E. Zakrajsek, 2009, “Credit Market Shocks and Economic Fluctuations: Evidence from Corporate Bond and Stock Markets,” NBER Working Paper 14863 (Cambridge, Massachusetts: National Bureau for Economic Research).
- Tornell, A., and F. Westermann, 2002, “Boom-Bust Cycles in Middle Income Countries: Facts and Explanation,” IMF Staff Papers, Vol. 49 Special Issue, pp. 111–53.
- Calvo, G. A., A. Izquierdo, and L. F. Mejia, 2004, “On the empirics of Sudden Stops: The Relevance of Balance Sheet Effects,” NBER Working Paper No. 10520 (Cambridge, Massachusetts: National Bureau for Economic Research).

### Monetary policy, exchange rates, and international transmission
- Bjornland, H. C., and J. I., Halvorsen, 2008, “How Does Monetary Policy Respond to Exchange Rate Movements? New International Evidence,” Working Paper 2008/15, Norges Bank.
- Canova, F., 2005, “The Transmission of U.S. Shocks to Latin America”, Journal of Applied Econometrics, Vol. 20, No. 2, pp. 229–51.
- Clarida, R., and J. Gali, 1994, “Sources of Real Exchange Rate Fluctuations: How Important are Nominal Shocks?” Carnegie-Rochester Conference on Public Policy, Vol. 41, pp. 1‒56.
- Farrant, K. and G. Peersman, 2006, “Is the Exchange Rate a Shock Absorber or Source of Shocks? New Empirical Evidence,” Journal of Money, Credit, and Banking, Vol. 38, pp. 939–62.
- Levy-Yeyati, E., and F. Sturzenegger, 2005, “Classifying Exchange Rate Regimes: Deeds vs. Words,” European Economic Review, Vol. 49, No. 6, pp. 1603–35.
- Peersman, G., 2005, “What Caused the Early Millennium Slowdown? Evidence Based on Vector Autoregressions,” Journal of Applied Econometrics, Vol. 20, pp. 185–207.
- Peersman, G., 2011, “The Relative Importance of Symmetric and Asymmetric Shocks: The Case of the United Kingdom and the Euro Area,” Oxford Bulletin of Economic and Statistics, Vol. 73, No. 1, pp. 104–18.
- Mountford, A., 2005, “Leaning into the Wind: A Structural VAR Investigation of U.K. Monetary Policy,” Oxford Bulletin of Economic and Statistics, Vol. 67, pp. 597–621.
- Bjornland, H. C., and J. I., Halvorsen, 2008, “How Does Monetary Policy Respond to Exchange Rate Movements? New International Evidence,” Working Paper 2008/15, Norges Bank.

### VAR methodology and identification techniques
- Sims, C.A., 1980, “Macroeconomics and Reality”, Econometrica, Vol. 48, No. 1, pp. 1-48.
- Faust, J., 1998, “The Robustness of Identified VAR Conclusions About Money,” Carnegie-Rochester Conference on Public Policy, Vol. 49, pp. 207–244.
- Uhlig, H., 2005, “What are the Effects of Monetary Policy on Output? Results from An Agnostic Identification Procedure,” Journal of Monetary Economics, Vol. 52, pp. 381–419.
- Fry, R., and A. Pagan, A., 2007, “Some Issues in Using Sign Restrictions for Identifying Structural VARs,” NCER Working Paper Series 14, National Centre for Econometric Research.
- Fry, R., and A. Pagan, A., 2010, “Sign Restrictions in Structural Vector Autoregressions: A Critical Review,” CAMA Working Papers 2010–22 (Australian National University, Centre for Applied Macroeconomic Analysis).

### Historical and broader empirical analyses
- Bernanke, Ben S., and Cara S. Lown, 1991, “The Credit Crunch,” Brookings Papers on Economic Activity, No. 2, pp. 205–47.
- Reinhart, C. and K. S. Rogoff, 2009, This Time is Different: Eight Centuries of Financial Folly (Princeton: Princeton Press).
- Woodford, M., 2003, Interest and Prices: Foundations of a Theory of Monetary Policy (Princeton: Princeton Press).

*Source: _wp1243 - REFERENCES*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2012/_wp1243.pdf_