## _cr1566

## Source details

**Canonical URL:** [_cr1566](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2015/_cr1566.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2015/_cr1566.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2015/_cr1566.pdf.json)

---

### IS THERE SCOPE FOR FURTHER DEDOLLARIZATION POLICIES?

- Stylized facts (Armenia)
  - In the third quarter of 2014, around 60 percent of total deposits and private sector credit was denominated in dollars, slightly below historical averages of 65 percent.
  - Deposit dollarization (average for 1995Q1 to 2014Q3) and credit dollarization (average for 1996Q4 to 2014Q3) have exhibited wide fluctuations, peaking at over 80 percent in the early 2000s and falling to below 40 percent in 2007–08.
  - Dollarization is closely linked to exchange rate developments and less linked to inflation developments; inflation was generally in single digits after 1999 and never exceeded 12 percent during that period.
  - Demand deposit dollarization was near 45 percent in the third quarter of 2014.
  - Definitions:
    - Deposit dollarization = fraction of deposits denominated in foreign currency to total deposits (broad money M2 definition).
    - Credit dollarization = credit to residents denominated in foreign currency as a share of total credit to residents.

- Armenia’s dedollarization strategy and components
  - Strategy mixes macroeconomic stabilization and specific dedollarization measures:
    - Macroeconomic stabilization achievements: reductions of inflation and the fiscal deficit; some narrowing of the external current account deficit, with remaining room for additional consolidation.
    - Dedollarization measures: restrictions on use of FX for certain transactions; prudential, supervisory, and crisis management measures.
  - Key prudential and regulatory elements:
    - Differential reserve requirements (RRs) and denomination of RRs in dram.
    - Measures to reduce FX liquidity risk and strengthen capital buffers.
    - Reduction in the rate of RRs for dram liabilities (recent).
    - Introduction of: (i) higher risk weights for the calculation of credit risk for FX loans; (ii) minimum liquidity ratios for FX liabilities; and (iii) improved monitoring of currency mismatches.
  - Rationale and international consistency:
    - Strategy aligns with international lessons: mix of macro stabilization, market-based dedollarization policies, and, where used, forced dedollarization policies.
    - Among market-based policies, prudential regulations (e.g., differential RRs) and supervisory measures are commonly preferred.

- Impact of dedollarization measures (international evidence and Armenia episodes)
  - International evidence:
    - Dollarization is persistent; dedollarization policies materialize gradually.
    - Reinhart, Rogoff, and Savastano (RRS) (2003): out of 85 countries (1980–2001), only four (Israel, Mexico, Pakistan and Poland) achieved significant reduction in deposit dollarization; only two (Israel and Poland) did so without severe distortions.
    - Inflation stabilization alone is not sufficient; a country’s inflation history is a better predictor of dollarization than current inflation.
    - Successful dedollarization cases combined disinflation programs with exchange rate anchors, availability of alternative indexed assets or high real interest rates, prudential policies, dollar substitutes, and favorable external financing.
  - Armenia-specific episodes:
    - 2005–07: deposit dollarization fell by almost 40 percent, driven by improved banking sector confidence, strong growth, 38 percent dram appreciation between end-2004 and end-2007, and restrictions on FX consumer loans.
    - Q1:2010 to Q3:2014: deposit dollarization fell from a peak of 73 percent to 58 percent, coinciding with improved macro conditions and prudential measures affecting interest rate spreads.
  - Prudential measures effects in Armenia:
    - Shift in currency denomination of RRs toward drams made more dollar resources available for lending while banks increased dram deposits to meet dram RRs.
    - Impact on credit dedollarization has been less effective than for deposit dedollarization.
    - Lending-deposit differential (dram minus FX) has not changed significantly in recent years.

- Assessment and scope for additional measures
  - Armenia’s strategy is comprehensive and consistent with international experience, reducing the need and scope for additional measures.
  - Suggested additional focus areas:
    - Reduce inflation volatility and external imbalances.
    - Use prudential regulations to increase foreign currency liquidity in the banking system.
    - Strengthen monitoring of currency mismatches.
  - Expectation on pace: further reductions in dollarization are likely to occur only gradually.

---

### Macroeconomic stabilization (findings and policy options)

- Findings
  - Inflation averaged 5.4 percent between 2003 and 2013, ranged from a low of -0.3 percent in 2005 to a high of 8.7 percent in 2010 with a standard deviation of 2.4 percent.
  - Much of the inflation volatility comes from international commodity prices.

- Policy options and considerations
  - Better communications on the sources of inflation volatility to improve the reliability of the dram as a store of value.
  - Creation of inflation-indexed instruments (note: could bring volatility risks into the banking system, e.g., risks of volatility of commodity prices).
  - Allow the exchange rate (ER) to move according to fundamentals to avoid one-side (depreciation) bets on the ER that would discourage private agents from rebalancing portfolios toward drams.

---

### Prudential regulation (vulnerabilities and recommendations)

- Vulnerabilities and findings
  - Strategy of relatively low RRs for FX liabilities, denomination of RRs in dram, and relatively low FX liquidity requirements may make the system vulnerable to FX liquidity shocks.
  - 2012 FSAP Update: while aggregate foreign currency liquidity was above minimum requirements, 14 out of 21 banks had foreign currency liquidity ratios below the aggregate ratios.
  - Fast growth of loans via FX credit cards since 2009 is noteworthy and may reflect the elimination of FX consumer loans in 2007.

- Policy recommendations
  - Continue and strengthen the CBA’s FX liquidity stress tests as an input for financial supervision.
  - Consider increasing FX liquidity requirements to address limited FX liquidity and reduce profitability of FX assets.
    - Implementation challenges: likely resistance from commercial banks due to impact on profitability.
    - Facilitating elements: reductions in dram RRs support bank profitability; implementation of Basel III guidelines implies an increase in overall liquidity ratios, presenting an opportunity to implement higher FX liquidity ratios.
  - Continue monitoring currency mismatches of bank borrowers and bank responses to regulatory changes.
  - Ensure banks internalize potential risks associated with new FX lending channels (e.g., FX credit cards).
  - Strengthen the monitoring template for assessment of unhedged borrowers (TA needed).

---

### Key policy measures, instruments, and calibration

- Reserve requirements and denomination
  - A reserve requirement of 2 percent applies to banks' dram liabilities and a reserve requirement of 12 percent applies to dollar liabilities.
  - All reserve requirements must be denominated in drams.

- Liquidity ratios (FX)
  - For assets denominated in foreign currency:
    - Liquidity ratio of "highly liquid assets to total assets" floor: 4 percent.
    - Ratio of "highly liquid assets to demand liabilities" floor: 10 percent.
  - At the balance sheet level:
    - Respective ratios: 15 percent and 60 percent.
  - Comment: Foreign currency liquidity ratios are less stringent than aggregate liquidity ratios and are unlikely to be very effective; this could lead to under-provision of FX liquidity at the bank level.

- Prudential limits and weights
  - Fifty percent additional risk weight for several categories of FX lending, including for loans classified as sub-standard and doubtful (September 2010).
  - Net open FX position limits: 7 percent of capital on individual currencies, 10 percent of capital on aggregate open FX positions.

- Deposit insurance
  - Bank deposits in domestic currency are covered up to 4 million AMD, while bank deposits in foreign currency are covered only up to 2 million AMD. All payments are made in AMD.

- Forced dedollarization measures
  - Prices of and payments for goods, services, wages, and investments in statutory and share capital legal entities are quoted and made in drams only (exceptions apply).
  - Consumer loans can only be extended in drams.

- Timeline of selected policy changes (exact phrasing preserved)
  - March 2010 (effective April 2010): Reserve requirements against funds attracted in foreign exchange are established as 12 percent, of which 3 percent in AMD and 9 percent in foreign exchange.
  - July 2010 (effective September 2010): Reserve requirements against funds attracted in foreign exchange are established as 12 percent, of which 6 percent in AMD and 6 percent in foreign exchange.
  - February 2011 (effective the same month): Reserve requirements against funds attracted in foreign exchange are established as 12 percent, of which 9 percent in AMD and 3 percent in foreign exchange.
  - November 2011 (effective December 2011): Reserve requirements against funds attracted in foreign exchange are established as 12 percent to be reserved in AMD.
  - April 2013 (effective June 2013): Reserve requirement for dram liabilities reduced from 8 percent to 4 percent.
  - February 2014 (effective March 2014): Reserve requirement for dram liabilities reduced from 4 percent to 2 percent.
  - June 2012 (effective January 2013): Minimum liquidity requirements for FX liabilities implemented.
  - June 2012: Monitoring of currency mismatches—banks to use a template to assess the extent of currency mismatches of banks' largest borrowers.

---

### Fiscal multipliers in Armenia — findings and implications

- Context and problem statement
  - Growth slowing and continuing budget underspending raised questions on fiscal policy’s role.
  - Last three years (relative to document): actual government expenditure fell short of budget targets, largely due to delays in large infrastructure projects.
  - In 2013, underexecution of the capital budget reached 48 percent (2.2 percent of GDP).
  - IMF-supported program aims to bolster growth through higher capital and social spending, supported by higher tax revenues raised in a growth-neutral way.

- Need for accompanying revenue measures
  - Increase in government expenditure needs to be accompanied by improved revenue collection to preserve macro stability.
  - The new IMF program under the EFF focuses on revenue enhancement measures and improving composition of spending.

- Estimation approaches and caveats
  - Fiscal multipliers are hard to estimate econometrically due to reverse causation.
  - Approaches: SVAR (when sufficient quarterly data exist); DSGE models; "bucket approach" when direct estimation not feasible.
  - Bucket approach is a rule-of-thumb calibrated largely on advanced-economy evidence and requires judgment for Armenia.

- Empirical reference values and Armenia-specific bucket result
  - Advanced economies: first-year spending multiplier averages about 0.8; revenue multiplier about 0.3 in normal times.
  - Bucket approach applied to Armenia suggests an intermediate case between “low” (0.1-0.3) and “medium” (0.4-0.6) multipliers in normal times, with caveats.

---

### Model-based (DGE) simulation results

- Model and shock
  - Model: small non-linear dynamic general equilibrium model calibrated to the Armenian economy (Cincibuch et al. (2011)).
  - Illustrative fiscal shock: permanent increase in government expenditure by 5 percent.

- Output response and implied multipliers
  - Output increases about 2¼ percent in the first year.
  - A 1 percentage point increase in government expenditure gives a cumulative increase in real GDP of 0.54 percentage points.
  - Translates into a spending multiplier in absolute terms close to 2.
  - Revenue multiplier: authors do not find any impact of revenue changes on output.

---

### Econometric (SVAR) estimation results

- SVAR setup
  - Identification following Blanchard and Perotti (2002): variables = net taxes, government spending, GDP.
  - Quarterly data: Q1 1999-Q4 2013.
  - Net taxes = tax revenue minus subsidies and social transfers; government spending = expenditure on goods and services and acquisition of non-financial assets.
  - Real terms, seasonally adjusted; estimated in levels (logarithms) with 4 lags, linear trend, and dummy for methodological change since Q1 2008.

- Identification assumptions
  - Tax elasticity to output assumed unit elasticity.
  - Spending elasticity to output assumed zero.

- SVAR multiplier estimates for Armenia
  - Output responds positively to a spending shock; small initially, peak after three quarters.
  - Peak/spending multiplier at peak: about 1.4.
  - First-year spending multiplier: about 0.89 (reported elsewhere as around 0.9).
  - Response to net taxes: virtually no impact on expenditure and a small negative effect on GDP initially that becomes insignificant from the first quarter on.

- Multipliers by expenditure type (capital spending)
  - Impact multiplier (capital spending): 0.21.
  - After one year (capital spending): 1.12.
  - Peak (capital spending, third quarter): 1.64.
  - For total expenditure on goods, services and capital: impact 0.08, after one year 0.89, peak 1.42.

- Comparison and synthesis
  - DGE model-based multiplier close to 2; SVAR first-year around 0.9; peak around 1.4.
  - Possible contributors to higher multipliers in Armenia: lower savings rate; financial constraints; relatively low public debt (under 50 percent of GDP); access to foreign budget financing.

---

### Policy implications and recommendations (fiscal)

- Fiscal policy effectiveness
  - Fiscal policy can have a significant impact on output in Armenia and can be used for short-term stimulus.
  - Channels: job creation, purchase of domestic inputs, elimination of bottlenecks via infrastructure projects.

- Importance of capital spending
  - High capital spending multipliers underscore importance of executing a well-designed and well-targeted public investment program.
  - Illustration: if the capital budget for 2013 had been executed as planned, this could have added an additional 2.5 percentage points to growth based on SVAR estimates for capital expenditures.

- Sectoral priorities and fiscal design
  - Prioritize infrastructure, healthcare and education where spending lags peers.
  - Any increased expenditure should be accompanied by improved revenue performance; moderate tax increases are not likely to have a significant negative effect on output per the results.

- Forecasting and policy design
  - Active consideration of fiscal multipliers would improve growth projections and design of policy packages.

_Italic: Source — Republic of Armenia, INTERNATIONAL MONETARY FUND (content unit: _cr1566)._

### References ________________________________________________________________________________ 14

### IS THERE SCOPE FOR FURTHER DEDOLLARIZATION POLICIES?

### A. Dollarization in Armenia: Stylized Facts
- In the third quarter of 2014, around 60 percent of total deposits and private sector credit was denominated in dollars, slightly below historical averages of 65 percent.
- Deposit dollarization (average for 1995Q1 to 2014Q3) and credit dollarization (average for 1996Q4 to 2014Q3) have exhibited wide fluctuations, peaking at over 80 percent in the early 2000s and falling to below 40 percent in 2007–08.
- Dollarization has been closely linked to exchange rate developments and less linked to inflation developments:
  - Examples: during 2005–07, when the dram was appreciating, dram deposits grew significantly faster than dollar deposits; in Q1:2009, there was a significant shift from dram deposits to dollar deposits amid sharp dram depreciation.
  - Inflation was generally in single digits after 1999 and never exceeded 12 percent during that period.
- Dollarization is more pronounced for time deposits but has also been high for demand deposits; demand deposit dollarization was near 45 percent in the third quarter of 2014.
- Definitions used in the note:
  - Deposit dollarization = fraction of deposits denominated in foreign currency to total deposits (broad money M2 definition).
  - Credit dollarization = credit to residents denominated in foreign currency as a share of total credit to residents.

### B. Dedollarization Policies (Armenia’s strategy and components)
- Armenia’s dedollarization strategy combines macroeconomic stabilization and specific dedollarization measures:
  - Macroeconomic stabilization achievements: reductions of inflation and the fiscal deficit; some narrowing of the external current account deficit, with remaining room for additional consolidation.
  - Dedollarization measures in place (see Table 1 and Table 2 in source): restrictions on use of FX for certain transactions; prudential, supervisory, and crisis management measures.
- Key prudential tools and regulatory elements used:
  - Differential reserve requirements (RRs) and use of the dram as the currency of denomination of RRs.
  - Measures to reduce FX liquidity risk and strengthen capital buffers.
  - More recent actions include a reduction in the rate of RRs for dram liabilities.
  - Introduction of: (i) higher risk weights for the calculation of credit risk for FX loans; (ii) minimum liquidity ratios for FX liabilities; and (iii) improved monitoring of currency mismatches.
- Rationale and international consistency:
  - Strategy aligns with international lessons: a mix of macro stabilization, market-based dedollarization policies, and, where used, forced dedollarization policies.
  - Among market-based policies, countries typically prefer prudential regulations (e.g., differential RRs for domestic and FX deposits) and supervisory measures.

### C. Impact of Dedollarization Measures
- International evidence on dedollarization:
  - Dollarization is persistent; dedollarization policies materialize gradually.
  - Vegh (2013) summary: households avoid currencies that fall over time and fluctuate wildly; restoring confidence requires a long track record of macro stability and consistent implementation of sound monetary and fiscal policies, and sometimes institutional reforms.
- Box 1: "Successful Dedollarization is the Exception, not the Norm"
  - Reinhart, Rogoff, and Savastano (RRS) (2003): out of 85 countries (1980–2001), only four (Israel, Mexico, Pakistan and Poland) achieved significant reduction in deposit dollarization; only two (Israel and Poland) did so without severe distortions in financial intermediation.
  - Dedollarization is gradual and subject to reversals (e.g., Armenia).
  - Inflation stabilization alone is not sufficient: a country’s inflation history tends to be a better predictor of dollarization than current inflation; changes in inflation do not have clear short-term impact on financial dedollarization.
  - Elements contributing to dedollarization in successful cases include strong disinflation programs with exchange rate anchors, availability of assets with alternative indexation or high real interest rates, prudential policies (necessary but risky), dollar substitutes (e.g., inflation-indexed instruments), and favorable external financing conditions.
- Armenia-specific impacts and episodes:
  - Two periods with notable progress:
    - 2005–07: deposit dollarization fell by almost 40 percent, driven mainly by macroeconomic reasons—improved banking sector confidence, strong growth, 38 percent dram appreciation between end-2004 and end-2007, and restrictions on FX consumer loans.
    - Since Q1:2010 to Q3:2014: deposit dollarization fell from a peak of 73 percent to 58 percent, coinciding with improved macro conditions (reduced fiscal deficit, some current account consolidation, relative exchange rate stability) and prudential measures affecting interest rate spreads.
  - Prudential measures effects:
    - Shift in currency denomination of RRs toward drams made more dollar resources available for lending while banks increased dram deposits to meet dram RRs.
    - Impact on credit dedollarization has been less effective than for deposit dedollarization.
    - It is unclear whether prudential measures have effectively modified intermediation spreads in FX: the lending-deposit differential (dram minus FX) has not changed significantly in recent years (see Figure 2).

### D. Is There Scope for Additional Dedollarization Policies in Armenia?
- Assessment:
  - Armenia’s dedollarization strategy is comprehensive and consistent with international experience, reducing the need and scope for additional measures.
  - The strategy includes: achieving relatively low inflation, pursuing a comprehensive strategy to reduce macroeconomic imbalances and vulnerabilities, and maintaining a broad set of dedollarization policies across key areas.
- Suggested additional focus areas (from analysis in the text):
  - Reduce inflation volatility and external imbalances.
  - Use prudential regulations to increase foreign currency liquidity in the banking system.
  - Strengthen monitoring of currency mismatches.
- Expectation on pace:
  - International experience suggests further reductions in dollarization are likely to occur only gradually.

*Republic of Armenia — International Monetary Fund (December 5, 2014).*

### 10.      Nonetheless, there appears to be room for action in a few areas:

### 10.      Nonetheless, there appears to be room for action in a few areas:

### Macroeconomic stabilization
- Finding: Reducing inflation volatility may help increase the credibility of the inflation targeting regime and lay a sounder foundation for dedollarization.
  - Inflation averaged 5.4 percent between 2003 and 2013, ranged from a low of -0.3 percent in 2005 to a high of 8.7 percent in 2010 with a standard deviation of 2.4 percent.
  - Much of the inflation volatility comes from international commodity prices.
- Policy options and considerations:
  - Better communications on the sources of inflation volatility to improve the reliability of the dram as a store of value.
  - Creation of inflation-indexed instruments (note: could bring volatility risks into the banking system, e.g., risks of volatility of commodity prices).
  - Allow the exchange rate (ER) to move according to fundamentals to avoid one-side (depreciation) bets on the ER that would discourage private agents from rebalancing portfolios toward drams.

### Prudential regulation
- Vulnerabilities and findings:
  - Armenia’s strategy of relatively low reserve requirements (RRs) for FX liabilities, denomination of RRs in dram, and relatively low FX liquidity requirements in FX may make the system vulnerable to FX liquidity shocks.
  - The 2012 FSAP Update found that while the system’s foreign currency liquidity was above the minimum requirements for aggregate liquidity, foreign currency liquidity was unevenly distributed: 14 out of 21 banks had foreign currency liquidity ratios below the aggregate ratios.
  - Fast growth of loans via FX credit cards since 2009 is noteworthy and may have been a response to the elimination of FX consumer loans in 2007.
- Policy recommendations:
  - Continue and strengthen the CBA’s FX liquidity stress tests as an important input for financial supervision.
  - Consider increasing FX liquidity requirements to address limited FX liquidity and contribute to dedollarization by reducing profitability of FX assets.
    - Implementation challenges: likely resistance from commercial banks due to impact on profitability.
    - Facilitating elements: recent reductions in dram RRs support bank profitability; implementation of Basel III guidelines implies an increase in overall liquidity ratios, presenting an opportunity to implement higher FX liquidity ratios.
  - Continue monitoring currency mismatches of bank borrowers and bank responses to regulatory changes.
  - Ensure banks internalize potential risks associated with new FX lending channels (e.g., FX credit cards).
  - Strengthen the monitoring template for assessment of unhedged borrowers (TA needed to improve operational use of this template).

### Key policy measures, instruments, and calibration (as described)
- Current RRs and denomination:
  - A reserve requirement of 2 percent applies to banks' dram liabilities and a reserve requirement of 12 percent applies to dollar liabilities.
  - All reserve requirements must be denominated in drams.
- Liquidity ratios (FX):
  - For assets denominated in foreign currency, the liquidity ratio of "highly liquid assets to total assets" has a floor of 4 percent, while the ratio of "highly liquid assets to demand liabilities" has a floor of 10 percent.
  - At the balance sheet level, the respective ratios are 15 percent and 60 percent.
  - Comment: Foreign currency liquidity ratios are less stringent than aggregate liquidity ratios and are unlikely to be very effective; this could lead to under-provision of FX liquidity at the bank level.
- Prudential limits and weights:
  - Higher risk weights for a variety of foreign currency assets, such as claims on individuals and legal entities (e.g., Fifty percent additional risk weight for several categories of FX lending, including for loans classified as sub-standard and doubtful, September 2010).
  - Net open FX position limits: 7 percent of capital on individual currencies, 10 percent of capital on aggregate open FX positions.
- Deposit insurance:
  - Bank deposits in domestic currency are covered up to 4 million AMD, while bank deposits in foreign currency are covered only up to 2 million AMD. All payments are made in AMD.
- Forced dedollarization measures:
  - Prices of and payments for goods, services, wages, and investments in statutory and share capital legal entities are quoted and made in drams only (exceptions apply).
  - Consumer loans can only be extended in drams.
- Timeline of recent policy changes (selected, exact phrasing preserved):
  - March 2010 (effective April 2010): Reserve requirements against funds attracted in foreign exchange are established as 12 percent, of which 3 percent in AMD and 9 percent in foreign exchange.
  - July 2010 (effective September 2010): Reserve requirements against funds attracted in foreign exchange are established as 12 percent, of which 6 percent in AMD and 6 percent in foreign exchange.
  - February 2011 (effective the same month): Reserve requirements against funds attracted in foreign exchange are established as 12 percent, of which 9 percent in AMD and 3 percent in foreign exchange.
  - November 2011 (effective December 2011): Reserve requirements against funds attracted in foreign exchange are established as 12 percent to be reserved in AMD.
  - April 2013 (effective June 2013): Reserve requirement for dram liabilities reduced from 8 percent to 4 percent.
  - February 2014 (effective March 2014): Reserve requirement for dram liabilities reduced from 4 percent to 2 percent.
  - June 2012 (effective January 2013): Minimum liquidity requirements for FX liabilities (as described above) implemented.
  - June 2012: Monitoring of currency mismatches—banks to use a template to assess the extent of currency mismatches of banks' largest borrowers.

### Fiscal multipliers in Armenia — findings and implications
- Context and problem statement:
  - With growth slowing and continuing budget underspending, questions about the role of fiscal policy in stimulating economic activity in Armenia have become increasingly relevant.
  - In the last three years (relative to the document), actual government expenditure fell short of budget targets, largely driven by delays in large infrastructure projects.
  - In 2013, underexecution of the capital budget reached 48 percent (2.2 percent of GDP).
  - The IMF-supported program aims to bolster growth through higher capital and social spending, supported by higher tax revenues raised in a growth-neutral way.
- Need for accompanying revenue measures:
  - An increase in government expenditure would need to be accompanied by improved revenue collection to preserve macroeconomic stability.
  - The new IMF program under the EFF focuses on revenue enhancement measures and improving composition of spending (including better remuneration for public sector employees and implementation of major investment projects).
- Estimation challenges and approaches:
  - Fiscal multipliers are hard to estimate econometrically due to reverse causation (government expenditure tends to be procyclical).
  - Two main approaches: structural vector autoregression model (SVAR) if sufficient quarterly time series exist; or dynamic stochastic general equilibrium (DSGE) models. A “bucket approach” (back-of-the-envelope) can be used when direct estimation is not feasible.
- Empirical reference values and Armenia-specific result from the bucket approach:
  - For advanced economies, the first-year spending multiplier averages about 0.8 whereas the revenue multiplier is about 0.3 in normal times.
  - Application of the “bucket approach” to Armenia suggests an intermediate case between “low” (0.1-0.3) and “medium” (0.4-0.6) multipliers in normal times.
  - Caveats: The bucket approach is a rule-of-thumb calibrated largely on empirical evidence from advanced economies and requires judgment for applicability to Armenia. An upward adjustment in multipliers could be justified if a large proportion of the economy is controlled by the government.

*Source: IMF staff report excerpts as provided in the content unit.*

### 7.      Model-based simulations suggest that the first-year government spending multiplier

### 7. Model-based simulations suggest that the first-year government spending multiplier is around 2

### Model-based (DGE) simulation results
- Model: small non-linear dynamic general equilibrium model calibrated to the Armenian economy (Cincibuch et al. (2011)).
- Illustrative fiscal shock: permanent increase in government expenditure by 5 percent.
- Output response: increase of output of about 2¼ percent in the first year.
- Implied relationships and multipliers:
  - A 1 percentage point increase in government expenditure gives rise to a cumulative increase in real GDP of 0.54 percentage points.
  - Taking into account the share of public expenditure in GDP, this relationship translates into a spending multiplier in absolute terms of close to 2.
- Revenue multiplier: authors do not find any impact of revenue changes on output.

### Econometric (SVAR) estimation approach and data
- Identification strategies discussed: recursive (Cholesky) approach and structural VAR (SVAR) relying on economic theory; Blanchard and Perotti (2002) identification highlighted.
- SVAR implemented for Armenia following Blanchard and Perotti (2002):
  - Variables: net taxes, government spending, GDP.
  - Data frequency and sample: quarterly data spanning the period Q1 1999-Q4 2013.
  - Definitions: net taxes = tax revenue minus subsidies and social transfers; government spending = expenditure on goods and services and acquisition of non-financial assets (capital expenditure).
  - Transformations and controls: all variables in real terms using the GDP deflator, seasonally adjusted; estimated in levels (logarithms) with 4 lags and inclusion of a linear trend and a dummy for methodological change since Q1 2008 related to GFS 2001 reporting standards.
- Identification assumptions used (in absence of Armenia-specific elasticity estimates):
  - Tax elasticity to output assumed to be unit elasticity.
  - Spending elasticity to output assumed to be zero.
  - Rationale: flat VAT and CIT rates and relatively small progressivity of PIT justify unit tax elasticity; expenditures do not vary with output because no unemployment insurance and family benefit eligibility not related to output.

### SVAR fiscal multiplier estimates for Armenia
- General SVAR results:
  - Output responds positively to a spending shock, small initially, peak effect after three quarters.
  - Peak/spending multiplier at peak: about 1.4 (Table 1).
  - First-year spending multiplier (after one year): about 0.89 (reported elsewhere as around 0.9).
  - Response to net taxes: virtually no impact on expenditure and a small negative effect on GDP initially that becomes insignificant from the first quarter on.
- Multipliers by expenditure type (capital spending SVAR):
  - Impact multiplier (capital spending): 0.21.
  - After one year (capital spending): 1.12.
  - Peak (capital spending, peak at third quarter): 1.64.
  - For total expenditure on goods, services and capital (comparison values): impact 0.08, after one year 0.89, peak 1.42 (peak in third quarter).

### Comparison with other countries and methods
- Bucket approach (applied to Armenia) yields a total score of 3 (borderline between “low” and “medium” multiplier country) and suggests multiplier ranges in “normal” times:
  - Low: 0.1–0.3
  - Medium: 0.4–0.6
  - High: 0.7–1.0
  - Adjustments: for expenditure-based shocks, values adjusted upwards by 50 percent; for tax shocks, values adjusted downward by 50 percent; further adjustments for business cycle position and monetary policy constraints indicated.
- Cross-country context (selected emerging and transition economies):
  - Spending multipliers range between 0.2 and 0.8; tax multipliers between 0.1 and 0.9 (Annex III in Batini et al. (2014) for Bulgaria, Croatia, Hungary, Poland, Romania).
  - Czech Republic: short-term expenditure multiplier 0.4–0.6; tax multiplier 0.1–0.3.
- Synthesis:
  - SVAR estimates and the DGE model-based calculations for Armenia indicate spending multipliers higher than the typical values reported for less developed economies (DGE-based close to 2; SVAR first-year around 0.9; peak around 1.4).
  - Possible contributors to higher multipliers in Armenia: lower savings rate; financial constraints (relatively weak financial intermediation and high cost of borrowing) limiting consumption smoothing; relatively low public debt (under 50 percent of GDP); access to foreign budget financing.

### Policy implications and recommendations
- Fiscal policy effectiveness:
  - Fiscal policy can have a significant impact on output in Armenia and can be effectively used to provide short-term stimulus to the economy.
  - Channels: creation of additional jobs, purchase of domestic inputs (goods and services), and elimination of bottlenecks and constraints (infrastructure projects) that facilitate transportation and trade.
- Importance of capital spending:
  - High capital spending multipliers highlight the importance of proper execution of a well-designed and well-targeted public investment program, especially during periods of growth slowdown.
  - Illustration: if the capital budget for 2013 had been executed as planned, this could have added an additional 2.5 percentage points to growth based on the SVAR estimates for capital expenditures.
- Sectoral priorities:
  - More active government involvement in infrastructure, healthcare and education (where spending falls behind levels in other transition countries) could bring benefits.
  - Any increased expenditure should be accompanied by improved revenue performance; moderate tax increases are not likely to have a significant negative effect on output per the results.
- Forecasting and policy design:
  - Active consideration of fiscal multipliers would be useful for improving the accuracy of growth projections and the design of policy packages.

_Italic: Source — Republic of Armenia, INTERNATIONAL MONETARY FUND (content unit: 7. Model-based simulations suggest that the first-year government spending multiplier is around 2)._

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2015/_cr1566.pdf_