## _wp05144 - Appendix Tables

## Source details

**Canonical URL:** [_wp05144 - Appendix Tables](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05144.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05144.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05144.pdf.json)

---

### Major findings and interpretation
- Ruble broad money growth and headline inflation movements:
  - Ruble broad money growth accelerated from 35 to 50 percent in 2003 while headline inflation fell from 15 to 12 percent.
  - Since 2004 ruble broad money growth slowed from almost 60 percent to around 25 percent by mid-2005 while inflation picked up.
- Core versus headline inflation:
  - Core inflation was roughly stable while headline inflation moved, indicating a substantial part of headline movements were temporary or seasonal.
- Role of foreign cash holdings:
  - Including foreign cash holdings (primarily U.S. dollars) in the money aggregate is crucial for understanding money–inflation dynamics in Russia.
  - Balance of payments data show foreign cash holdings outside the banking system roughly constant at around $35 billion from 1998 through early 2003, falling by around $10 billion since then (from $36.9 billion in Q1 2003 to $27.1 billion in Q1 2005, with a temporary pickup in Q3 2004).
  - Effective broad money (M2 plus foreign currency deposits plus foreign cash holdings) growth has been much more stable than ruble broad money (M2) growth.

### Inflation dynamics and decomposition
- Long-run attribution (loglinear, restricted linear homogeneity α+β+γ=1):
  - 50 percent to nominal effective depreciation,
  - 40 percent to growth in unit labor costs,
  - 10 percent to utility price growth.
- Cointegration and persistence:
  - Johansen cointegration tests (April 1996–January 2004) generally support a unique cointegrating vector for the long-run inflation relation.
  - Deviations from the long-run relationship can be persistent—taking up to 12 months to be corrected.
- Short-run dynamics and money:
  - An excess supply of effective broad money appears inflationary, while excess supplies of narrower aggregates do not.
  - Changes in effective broad money growth have the strongest and most persistent effect on short-run inflation.
  - Estimated pass-through from nominal effective depreciation is relatively precise and close to the 0.5–0.6 pass-through estimate found elsewhere.

### Analytical framework and methodology
- Framework:
  - Combines markup theory of inflation (long-run determinants) and the monetary theory of inflation (short-run dynamics).
- Key variables in the long-run loglinear inflation equation:
  - Unit labor costs (ULC) = industrial wage bill / real industrial output.
  - Import price growth proxied by nominal effective exchange rate depreciation (neer).
  - Utility price growth proxied by CPI “paid services” component.
- Estimation approach (two-stage):
  1. Estimate long-run markup equation for inflation (cointegration approach).
  2. Estimate long-run money demand equation (cointegration approach).
  3. Combine long-run relationships and include equilibrium correction terms to determine short-run inflation dynamics.
- Data and stationarity:
  - Monthly data used for April 1996–January 2004.
  - Unit root tests indicate variables are nonstationary in levels but monthly changes in year-on-year growth rates are stationary.

### Long-run money demand: aggregates and key results
- Monetary aggregates defined:
  - RCC = Ruble Currency in Circulation.
  - RNM = RCC + RDD (ruble demand deposits).
  - RBM = RCC + RDD + RTD (ruble time and savings deposits).
  - BM = RCC + RDD + RTD + FCD (foreign currency deposits).
  - EBM = RCC + RDD + RTD + FCC (foreign currency in circulation estimate).
- Cointegration results:
  - Money demand for effective broad money (EBM) is the most stable (unique cointegrating vector), even at fixed accounting exchange rates.
- Selected unrestricted long-run coefficients (April 1996–January 2004; ** denotes 1 percent significance):
  - Ruble currency in circulation: Industrial Production 2.087 **, Deposit Rate −0.009 **, Depreciation −0.005 **.
  - Ruble narrow money: Industrial Production 1.319 **, Deposit Rate −0.010 **, Depreciation −0.004 **.
  - Ruble broad money: Industrial Production 3.159 **, Deposit Rate −0.004, Depreciation −0.004 **.
  - Broad money: Industrial Production 1.853 **, Deposit Rate −0.006 **, Depreciation −0.003 **.
  - Effective broad money: Industrial Production 1.216 **, Deposit Rate −0.008 **, Depreciation 0.001.
- Interpretation of coefficients:
  - Income elasticity (industrial production) is significant and generally exceeds unity across aggregates.
  - Deposit rate semi-elasticities range between −0.006 and −0.01.
  - Depreciation semi-elasticities for aggregates excluding foreign currency in circulation are approximately −0.003 to −0.005.

### Currency substitution, valuation effects, and 2003 interpretation
- Dependence on depreciation:
  - Measures excluding foreign currency in circulation are strongly negatively dependent on nominal depreciation, consistent with foreign cash as a substitute for domestic money.
  - Ruble broad money depreciation semi-elasticity was estimated at 0.4; an alternative estimate of 0.3 percent is not significantly different from 0.4.
- Implication for 2003:
  - Real ruble broad money grew by 30.6 percent during 2003.
  - Nominal ruble depreciation turned from 5.6 percent in 2002 to ruble appreciation of 7.8 percent in 2003.
  - The appreciation of 2003 contributed about 5½ percentage points to the 30½ percent growth in real money demand (derived from ruble broad money semi-elasticity result).
- EBM valuation versus volume:
  - EBM responds to exchange rate changes via valuation and volume effects; valuation effects dominated immediately after August 1998 but volume effects often larger before and after the crisis.
  - During 2003, nominal appreciation contributed negatively to EBM growth by encouraging declines in foreign currency deposits and cash holdings.

### Short-run inflation dynamics: ECM results and EBM effects
- Model structure:
  - ECM for ΔΔp includes lagged differences, EC∆p_{t-1}, ECm_{t-1}, and lagged ΔΔm_{t-i}.
  - b9 captures short-run correction of inflation to goods-market equilibrium deviations; b10 captures correction to money-market deviations.
- Best-fitting aggregate:
  - EBM yields best fit: lowest sigma, highest R-squared, maximum log-likelihood, smallest RSS, and most favorable residual-test p-values.
- Short-run estimated effects (selected):
  - An excess money supply of 1 percent in EBM (ECm_{t-1} = 1%) translates into a 3.7 percent acceleration in inflation (ECm_{t-1} (EBM) = 3.709 *).
  - Total size of the impact of a one-percent acceleration in EBM growth: 0.255.
  - Second largest total impact: ruble currency in circulation = 0.144.
  - Total impact of changes in excess narrow or broad ruble money is negative.
- Dynamics and persistence:
  - Changes in EBM growth have the strongest and most persistent effect on short-run inflation; EBM effects continue after four and five months.
  - Speed of adjustment of inflation to long-run equilibrium (EC∆p_{t-1} coefficients):
    - Range from −0.168 to −0.085 depending on monetary aggregate.
    - In the EBM equation EC∆p_{t-1} = −0.085 (significant at the 12 percent level).
    - Interpretation: 8.5 percent to 17 percent of a 1 percentage point deviation is corrected each month → adjustment time about 6 to 12 months.

### Quantitative model diagnostics (selected figures from Table 4)
- Total effect of ΔΔm (sum of lagged money-change coefficients) across aggregates: 0.144, −0.101, −0.085, 0.067, 0.255.
- No. of observations: 85.
- No. of parameters: 22.
- Sigma across equations: 1.56, 1.59, 1.59, 1.34, 1.13.
- R-squared across equations: 0.94, 0.94, 0.94, 0.95, 0.97.
- Log-likelihoods: −145.85, −147.30, −147.27, −132.49, −118.61.
- RSS: 153.93, 159.28, 159.18, 112.42, 81.09.
- F(21,63) statistics: 45.36 ***, 43.73 ***, 43.76 ***, 63.21 ***, 88.79 ***.
- Example short-run coefficients (EBM column):
  - EC∆p_{t-1}: −0.085.
  - ECm_{t-1} (EBM): 3.709 *.
  - ΔΔneer_{t-1} (EBM): 0.165 ***.
  - ΔΔp_{ut,t-1} (EBM): 0.394 ***.

### Valuation and volume decomposition (Appendix II)
- Components of EBM change:
  - Contribution of RBM growth,
  - Contribution of exchange rate changes (valuation effect),
  - Contribution of changes in volume of foreign currency assets (volume effect).
- Empirical patterns:
  - Valuation effects dominated in the year after the August 1998 crisis (fourfold ruble depreciation) and largely drove the EBM–inflation correlation in the four quarters after 1998.
  - Before and after the crisis the volume effect was often larger than the valuation effect; both were generally dominated by changes in RBM.
  - Evidence of “dollarization hysteresis”: post-1998 valuation increase in foreign assets was not quickly reversed, implying persistence.

### Estimating foreign currency in circulation: data, assumptions, uncertainty (Appendix I)
- BEA-based methodology:
  - Uses CBR estimates of net foreign currency sales by authorized banks’ exchange offices and net withdrawals from foreign currency deposits, adjusted for travel and shuttle trade outflows.
- Key assumptions and limitations:
  - Stock assumption: foreign cash holdings before December 1991 assumed zero.
  - For pre-1996 periods several ad hoc assumptions made where monthly CBR and BOP series were missing.
  - Travel expenditures assumed negligible in 1991 and grown at a constant rate until first BOP estimate in 1994.
- Comparison with alternative estimates:
  - CMIR-based estimate: U.S. dollar holdings estimated at $63 billion in November 1998.
  - Rimashevskaya (1998) survey (October 1996): $56 billion; CMIR estimate: $43 billion; BEA estimate: $8 billion.
  - American Express survey (end-2002): $13.5 billion; CBR: $39 billion for same period; BEA contact-adjusted American Express ≈ $18 billion.
- Despite uncertainty in level (stock), flows are estimated more reliably; adding a constant $35 billion to BEA series for April 1996–November 1998 did not significantly change regression results.

### Conclusions and policy relevance
- Main conclusions:
  - Accounting for foreign cash holdings explains part of the instability of money demand in Russia.
  - Inflation and money demand in dollarized economies cannot be well understood without including foreign cash holdings.
  - Long-run decomposition: nominal effective depreciation ≈ 50 percent of inflation, unit labor cost growth ≈ 40 percent, utility price growth ≈ 10 percent.
  - The inflation equation respects linear homogeneity (marginal effects sum to unity).
  - Large pass-through from depreciation to inflation implies further nominal appreciation may help reduce inflation.
- Policy-relevant implications:
  - Inflation does not significantly respond to excess supplies of aggregates that exclude foreign cash holdings, but does respond significantly to excess supply of EBM.
  - For each 1 percent by which EBM supply exceeds demand, inflation accelerates by 3.7 percent (short-run estimate).
  - An acceleration in domestic money growth need not be inflationary if it reflects dedollarization and monetization; conversely, slowdown in domestic money growth need not imply lower inflation if it reflects slowdown in dedollarization or renewed dollarization.
- Caveats and recommended improvements:
  - Foreign cash holdings estimates are uncertain and should be compared with other estimates.
  - Short-term EBM impact on inflation may partly reflect valuation effects.
  - Cointegration interpretation is limited by short sample and structural break around August 1998.
  - Suggested model improvements include correcting ULC for informal wages, Granger causality/weak exogeneity tests between wages and prices, alternative inflation and utility-cost measures, modeling exchange rate expectations, including financial innovation and banking confidence measures, separate demand functions for EBM components, and estimating additional long-run relationships in a simultaneous VAR.

*Source: _wp05144 - Appendix Tables (April 1996–January 2004 analyses and appendices as presented in the supplied content).*

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

### _wp05144 - References..............................................................................................................

### Figures
- 1. Core and Headline CPI and Ruble Broad Money .................................................................................................. 3
- 2. Effective Broad Money Components................................................................................................................... 3
- 3. Monetary Aggregates and Exchange Rate ......................................................................................................... 4
- 4. Headline CPI and Ruble-Dollar Exchange Rate ................................................................................................ 7
- 5. Headline CPI and Monetary Aggregates .......................................................................................................... 7

### Appendix Figures
- A1. Estimates of Foreign Cash Holdings in Russia ............................................................................................. 23
- A2. Inflation and Contributions to Effective Broad Money Growth ................................................................. 24

### Tables
- 1. Selected Transition Economies: Money Growth and Inflation, 2003............................................................ 6
- 2. Estimated Long-Run Cointegration Equation for Inflation, April 1996–January 2004 ............................. 11
- 3. Estimated Long-Run Money Demand Equations, April 1996–January 2004 .......................................... 15
- 4. Estimated Short-Run Inflation Equations, January 1997–January 2004 .................................................. 18

*https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05144.pdf*

### Appendix Tables

### _wp05144 - Appendix Tables

### Major findings and interpretation
- Ruble broad money growth and headline inflation showed divergent movements in the early 2000s: ruble broad money growth accelerated from 35 to 50 percent in 2003 while headline inflation fell from 15 to 12 percent; since 2004 ruble broad money growth slowed from almost 60 percent to around 25 percent by mid-2005 while inflation picked up.
- Core inflation was roughly stable while headline inflation moved, indicating a substantial part of headline movements were temporary or seasonal.
- Including foreign cash holdings (primarily U.S. dollars) in the money aggregate is crucial for understanding money–inflation dynamics in Russia.
- Following rapid dollarization until end-1999, Russia experienced dedollarization in recent years; balance of payments data show foreign cash holdings outside the banking system roughly constant at around $35 billion from 1998 through early 2003, falling by around $10 billion since then (from $36.9 billion in Q1 2003 to $27.1 billion in Q1 2005, with a temporary pickup in Q3 2004).
- Effective broad money (M2 plus foreign currency deposits plus foreign cash holdings) growth has been much more stable than ruble broad money (M2) growth.

### Inflation dynamics and decomposition
- Using a loglinear long-run specification, estimated long-run inflation dependence (restricted linear homogeneity α+β+γ=1) attributes roughly:
  - 50 percent to nominal effective depreciation,
  - 40 percent to growth in unit labor costs,
  - 10 percent to utility price growth.
- Johansen cointegration tests (April 1996–January 2004) generally support a unique cointegrating vector for the long-run inflation relation.
- Deviations from the long-run relationship can be persistent—taking up to 12 months to be corrected.
- Short-run inflation dynamics depend importantly on effective broad money:
  - An excess supply of effective broad money appears inflationary, while excess supplies of narrower aggregates do not.
  - Changes in effective broad money growth have the strongest and most persistent effect on short-run inflation.
- The estimated pass-through from nominal effective depreciation is relatively precise and close to the 0.5–0.6 pass-through estimate found elsewhere.

### Analytical framework and methodology
- The paper combines the markup theory of inflation (long-run determinants) and the monetary theory of inflation (short-run dynamics).
- Long-run inflation equation (loglinear) uses:
  - unit labor costs (ULC) calculated as industrial wage bill divided by real industrial output,
  - import price growth proxied by nominal effective exchange rate depreciation (neer),
  - utility price growth proxied by the CPI “paid services” component.
- Two-stage estimation approach:
  1. Estimate long-run markup equation for inflation (cointegration approach).
  2. Estimate long-run money demand equation (cointegration approach).
  3. Combine long-run relationships and include equilibrium correction terms to determine short-run inflation dynamics.
- Monthly data used for April 1996–January 2004; unit root tests indicate variables are nonstationary in levels but monthly changes in year-on-year growth rates are stationary.

### Long-run money demand: aggregates and results
- Monetary aggregates defined:
  - Ruble Currency in Circulation (RCC) = RCC
  - Ruble Narrow Money (RNM) = RCC + RDD (ruble demand deposits)
  - Ruble Broad Money (RBM) = RCC + RDD + RTD (ruble time and savings deposits)
  - Broad Money (BM) = RCC + RDD + RTD + FCD (foreign currency deposits)
  - Effective Broad Money (EBM) = RCC + RDD + RTD + FCC (foreign currency in circulation estimate)
- Cointegration tests indicate money demand for effective broad money is the most stable (unique cointegrating vector), even when calculated at fixed accounting exchange rates to exclude valuation effects.
- Estimated long-run money demand coefficients (unrestricted, April 1996–January 2004) show anticipated signs:
  - Industrial production coefficient (income elasticity) is significant and generally exceeds unity across aggregates.
  - Deposit rate semi-elasticities range between −0.006 and −0.01 (i.e., an increase in the deposit rate by 1 percentage point leads to a decrease in money demand by 0.6 percent to 1 percent).
  - Depreciation semi-elasticities for aggregates excluding foreign currency in circulation are approximately −0.003 to −0.005 (interpreted in text as roughly 0.4 percent impact in many cases), indicating strong negative dependence on nominal depreciation.
- Table 3 estimated coefficients (unrestricted):
  - Ruble currency in circulation: Industrial Production 2.087 **, Deposit Rate −0.009 **, Depreciation −0.005 **
  - Ruble narrow money: Industrial Production 1.319 **, Deposit Rate −0.010 **, Depreciation −0.004 **
  - Ruble broad money: Industrial Production 3.159 **, Deposit Rate −0.004, Depreciation −0.004 **
  - Broad money: Industrial Production 1.853 **, Deposit Rate −0.006 **, Depreciation −0.003 **
  - Effective broad money: Industrial Production 1.216 **, Deposit Rate −0.008 **, Depreciation 0.001
- Notes on significance: ** denotes significance at the 1 percent level.

### Interpretation of currency substitution and valuation effects
- All measures of money demand that exclude foreign currency in circulation are strongly negatively dependent on the nominal depreciation rate, consistent with foreign cash being an important substitute for domestic money.
- For ruble broad money, approximately one sixth of the growth in real ruble broad money in 2003 was accounted for by ruble appreciation (contextual numbers: real ruble broad money grew by 30.6 percent during 2003; nominal ruble depreciation turned from 5.6 percent in 2002 to ruble appreciation of 7.8 percent in 2003).
- Effective broad money responds to exchange rate changes both via valuation effects and via volume effects; Appendix II discusses valuation versus volume decomposition.

### Robustness, data limitations, and cautions
- Foreign currency in circulation estimate (FCC) is subject to considerable uncertainty (Appendix I).
- Short-term impact of effective broad money on inflation may be overestimated because of valuation effects (Appendix II).
- Cointegration results should be treated cautiously because of relatively short time series and use of monthly data with dummies for the August 1998 crisis, which may not fully remove structural breaks.
- Unit root tests have weak power in short samples; standard tests could not reject nonstationarity of variables in differences at the 1 percent level, but Johansen cointegration tests still generally indicate unique cointegrating vectors.
- The two-stage estimation method is commonly used when data length and quality are limited; the approach requires further testing with other data sources, additional variables, and different econometric specifications.

### Comparative observation for transition economies (2003)
- Selected 2003 average annual percent growth (Inflation, M2, Broad Money):
  - Azerbaijan: Inflation 2.2, M2 28.0, Broad Money 30.0
  - Belarus: Inflation 28.4, M2 71.1, Broad Money 56.8
  - Bulgaria: Inflation 2.3, M2 20.3, Broad Money 16.8
  - Kazakhstan: Inflation 6.4, M2 55.1, Broad Money 29.2
  - Kyrgyz Republic: Inflation 2.7, M2 33.5, Broad Money 33.4
  - Russia: Inflation 13.7, M2 51.6, Broad Money 39.7
  - Ukraine: Inflation 5.2, M2 43.5, Broad Money 47.2

- Several transition economies showed rapid money growth in 2003 while inflation remained relatively low, suggesting the possible relevance of including foreign cash holdings in money definitions for these countries as well.

*Source: _wp05144 - Appendix Tables (April 1996–January 2004 analyses and appendices as presented in the supplied content).*

### 2003. The predicted growth in real ruble broad money from this turnaround in ruble

### _wp05144 - 2003. The predicted growth in real ruble broad money from this turnaround in ruble

### Findings
- 2003. The predicted growth in real ruble broad money from this turnaround in ruble (continued...)
- money appears to be somewhat less sensitive to the depreciation rate with a semi-elasticity of

### Notes
- Text provided is fragmentary and ends mid-sentence at "semi-elasticity of".

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05144.pdf*

### 0.3 percent, this semi-elasticity is not significantly different from 0.4. As predicted, effective

### _wp05144 - 0.3 percent, this semi-elasticity is not significantly different from 0.4. As predicted, effective

### Money demand: stability, semi-elasticities, and the role of foreign cash holdings
- Effective broad money (EBM), which includes foreign currency deposits and an estimate of foreign currency in circulation, constitutes the most stable money demand function; ruble broad money is the most unstable.
- Coefficient estimates for the EBM demand equation are the most precisely estimated (narrowest confidence intervals); coefficient estimates for ruble broad money are the least precise and the coefficient on the deposit rate is insignificant for ruble broad money.
- Recursive estimates show:
  - Confidence intervals for EBM coefficient estimates remain stable over time.
  - Confidence intervals widen over time for money aggregates that exclude foreign currency.
- Long-run depreciation semi-elasticities:
  - Ruble broad money depreciation semi-elasticity was estimated at 0.4.
  - An alternative semi-elasticity reported as 0.3 percent is not significantly different from 0.4 (textual statement).
- Implications of the 2003 appreciation:
  - The appreciation of 2003 contributed about 5½ percentage points to the 30½ percent growth in real money demand (derived from ruble broad money semi-elasticity result).

### Short-run inflation dynamics: equilibrium-correction model (ECM) and EBM
- Model structure:
  - An equilibrium-correction model for ΔΔp (change in inflation) includes lagged differences of explanatory variables, an equilibrium correction term for inflation (EC∆p_{t-1}), an equilibrium correction term for money (ECm_{t-1}), and lagged changes in money growth (ΔΔm_{t-i}).
  - Coefficient b9 captures short-run correction of inflation to deviations from the long-run goods-market equilibrium; b10 captures short-run correction to deviations from the long-run money-market equilibrium.
  - Terms Σ_{i=6}^{10} b_{m+i} ΔΔm_{t-i} allow for direct effects of acceleration in money growth on acceleration in inflation.
- Best-fitting monetary aggregate:
  - The ECM estimated with effective broad money (EBM) gives the best fit: lowest sigma, highest R-squared, maximum log-likelihood, smallest RSS, and most favorable residual-test p-values (normality, autocorrelation, heteroskedasticity).
- Short-run responses:
  - Inflation does not significantly respond to excess supply of monetary aggregates that exclude foreign currency in circulation.
  - Inflation does significantly respond to excess supply of EBM.
  - Estimated EBM excess-money effect: an excess money supply of one percent (ECm_{t-1} = 1%) translates into a 3.7 percent acceleration in inflation (example: if inflation was initially 10 percent, it will be 10.37 percent the next month).
- Dynamics and persistence:
  - Changes in EBM growth have the strongest and most persistent effect on short-run inflation.
  - For most monetary aggregates the effect of an acceleration in money growth occurs mainly after three months; for EBM growth the effect continues after four and five months (more persistent).
  - Total size of the impact of a one-percent acceleration in effective broad money growth: 0.255 (interpreted as an acceleration of inflation by approximately ¼ percent).
  - Second largest total impact: ruble currency in circulation = 0.144.
  - Total impact of changes in excess narrow or broad ruble money is negative (consistent with instability of those money demand functions).
- Speed of adjustment of inflation to long-run equilibrium:
  - Coefficient on lagged equilibrium correction term for inflation (EC∆p_{t-1}) ranges from -0.168 to -0.085 depending on monetary aggregate.
  - In the EBM equation the coefficient equals -0.085 (significant at the 12 percent level).
  - Interpretation: if inflation exceeds its long-run equilibrium by 1 percentage point, 8.5 percent to 17 percent of this deviation is adjusted for every month → it takes about 6 to 12 months for inflation to return to its long-run equilibrium.
  - Conclusion: shocks to inflation are very persistent under the EBM description.

### Quantitative estimates and model diagnostics (selected table figures)
- Total effect of ΔΔm (sum of lagged money-change coefficients) across aggregates: 0.144, -0.101, -0.085, 0.067, 0.255 (values reported in Table 4).
- No. of observations: 85 (for all reported regressions).
- No. of parameters: 22 (for all reported regressions).
- Sigma values across equations: 1.56, 1.59, 1.59, 1.34, 1.13.
- R-squared across equations: 0.94, 0.94, 0.94, 0.95, 0.97.
- Log-likelihoods: -145.85, -147.30, -147.27, -132.49, -118.61.
- RSS (residual sum of squares): 153.93, 159.28, 159.18, 112.42, 81.09.
- F(21,63) statistics (significance): 45.36 ***, 43.73 ***, 43.76 ***, 63.21 ***, 88.79 *** (*** indicates significance at the 1-percent level).
- Diagnostic tests (examples from Table 4):
  - DW: 1.71, 1.92, 1.86, 2.26, 1.92.
  - AR 1-6 (F-test) p-values: (0.28), (0.46), (0.43), (0.63), (0.99).
  - ARCH 1-6 (F-test) p-values: (0.93), (0.80), (0.56), (0.21), (1.00).
  - Normality test (χ2-test) p-values: (0.59), (0.75), (0.68), (0.37), (0.71).
  - Heteroskedasticity (F-test) p-values: (0.08), (0.02), (0.15), (0.44), (0.40).
  - RESET test (F-test) p-values: (0.63), (0.40), (0.43), (0.83), (0.88).
- Selected short-run coefficient examples (EBM column shown where available):
  - EC∆p_{t-1}: -0.085 (EBM; significant at 12 percent level as text notes).
  - ECm_{t-1} (EBM): 3.709 * (significance indicated at 10-percent level in table notation).
  - Selected ΔΔneer_{t-1} coefficients: 0.165 *** (EBM; significant at 1-percent level).
  - Selected ΔΔp_{ut,t-1} coefficients: 0.394 *** (EBM; significant at 1-percent level).

### Valuation and volume effects on EBM growth (Appendix II findings)
- Decomposition of changes in EBM into:
  - Contribution of ruble broad money growth (RBM growth).
  - Contribution of exchange rate changes (valuation effect).
  - Contribution of changes in the volume of foreign currency assets (volume effect).
- Empirical findings:
  - Valuation effects dominated in the year following the August 1998 crisis (fourfold ruble depreciation); this valuation effect largely drove the correlation between EBM growth and inflation in the four quarters after the 1998 crisis.
  - Before and after the crisis the volume effect was often larger than the valuation effect; both were generally dominated by changes in ruble broad money.
  - During 2003, nominal appreciation of the ruble against the U.S. dollar encouraged declines in foreign currency deposits and cash holdings and thus had a negative contribution to EBM growth.
- Rationale for using EBM at actual exchange rates:
  - Sudden depreciation converts to an apparent excess money supply in the short run (valuation effect), which can be inflationary; sudden appreciation implies excess money demand, which can be disinflationary.
  - Short-run valuation effects can become longer-run volume effects due to “dollarization hysteresis” (evidence for Russia). Example: the post-1998 valuation increase in foreign assets was not quickly reversed, implying persistence.

### Estimating foreign currency in circulation: data, assumptions, and uncertainty (Appendix I)
- BEA estimate methodology:
  - Based on CBR estimates of net foreign currency sales by authorized banks’ exchange offices and net withdrawals from foreign currency deposits.
  - Adjusted by the Bureau of Economic Analysis (BEA) in Moscow for outflows related to travel and shuttle trade (shuttle trade estimated via BOP “other corrections for imports”; travel via BOP “travel services”).
- Key assumptions and data limitations:
  - Stock assumption: foreign cash holdings before December 1991 assumed zero (BEA assumption adopted by authors).
  - For years prior to 1996, several ad hoc assumptions due to missing CBR and BOP monthly series (e.g., monthly net foreign currency sales before 1996 assumed equal to 1/12 of annual estimates).
  - Travel expenditures assumed negligible in 1991 and grew at a constant rate until first BOP estimate in 1994.
- Comparison with alternative estimates:
  - CMIR-based estimate (U.S. Customs CMIR flows) available through November 1998: U.S. dollar holdings estimated at $63 billion in November 1998.
  - Survey estimate by Rimashevskaya (1998) for October 1996: $56 billion; CMIR estimate for that period: $43 billion; BEA estimate for that period: $8 billion.
  - American Express survey described by Semenov (2003): foreign cash holdings at end-2002 estimated at $13.5 billion; CBR estimates $39 billion for same period. Adjustments for individuals per household raise American Express survey to $18 billion (BEA contact estimate).
- Despite uncertainty in level (stock), less uncertainty exists about flows; across April 1996–November 1998 the difference between CMIR and BEA series is almost constant at around $35 billion; adding $35 billion to BEA estimates did not significantly change regression results.

### Conclusions and policy-relevant implications (from IV. CONCLUSION)
- Main conclusions:
  - Accounting for foreign cash holdings explains part of the instability of money demand in Russia (and more generally in dollarized transition economies).
  - Inflation and money demand in dollarized economies cannot be well understood without including foreign cash holdings.
  - In the long run, nominal effective depreciation accounts for roughly 50 percent of inflation, unit labor cost growth accounts for about 40 percent, and utility price growth for about 10 percent.
  - The inflation equation is consistent with linear homogeneity (marginal effects of input cost inflation sum to unity).
  - Large pass-through from depreciation to inflation implies that further nominal appreciation may help reduce inflation.
- Policy-relevant points:
  - Inflation appears not to significantly respond to excess supplies of monetary aggregates that exclude foreign cash holdings, but does respond significantly to excess supply of EBM.
  - For each 1 percent by which EBM supply exceeds demand, inflation accelerates by 3.7 percent (short-run estimate).
  - An acceleration in domestic money growth need not be inflationary to the extent it reflects dedollarization and monetization; conversely, slowdown in domestic money growth need not imply lower inflation if it reflects slowdown in dedollarization or renewed dollarization.
- Caveats and suggested improvements:
  - Estimates of foreign cash holdings are subject to considerable uncertainty and should be compared with other estimates.
  - Short-term impact of EBM on inflation may partly reflect valuation effects despite attempts to correct for exchange rate impacts.
  - Interpreting cointegrating vectors as true “long-run” equations is difficult given short time series and structural break around August 1998.
  - Suggested model improvements include: correcting unit labor cost estimates for informal sector wages; Granger causality or weak exogeneity tests between wages and prices; alternative inflation measures (core or goods inflation); alternative measures of utility costs; modeling exchange rate expectations (ratchet variables); including measures of financial innovation and banking confidence; estimating separate demand functions for EBM components; estimating additional long-run relationships (interest rates and exchange rates) ideally in a simultaneous VAR; testing out-of-sample forecasting usefulness.

*IMF Working Paper content from the supplied PDF extract.*

### References

### _wp05144 - References

### Money demand, currency substitution, and dollarization
- Bahmani-Oskooee, Mohsen, and Michael P. Barry, 2000, “Stability of the Demand for Money in an Unstable Country: Russia,” Journal of Post-Keynesian Economics, Vol. 22, No. 4, pp. 619–29.
- Banerji, Angana, 2002, “Money Demand,” Russian Federation: Selected Issues and Statistical Appendix, IMF Staff Country Report No. 02/75 (Washington: International Monetary Fund).
- Choudhry, Taufiq, 1998, “Another Visit to the Cagan Model of Money Demand: The Latest Russian Experience,” Journal of International Money and Finance, Vol. 17, No. 2, pp. 355-76.
- Feige, Edgar L., 2002, “Currency Substitution, Unofficial Dollarization, and Estimates of Foreign Currency Held Abroad: The Case of Croatia,” in Financial Policies in Emerging Markets, ed. by Mario Blejer and Marko Skreb (Cambridge, Massachusetts: MIT Press).
- Mongardini, Joannes, and Johannes Mueller, 2000, “Ratchet Effects in Currency Substitution: An Application to the Kyrgyz Republic,” IMF Staff Papers, Vol. 47, No. 2, pp. 218–37.
- Oomes, Nienke, 2003, “Network Externalities and Dollarization Hysteresis: The Case of Russia,” IMF Working Paper 03/96 (Washington: International Monetary Fund).
- Semenov, Sergey, 2003, “Dedollarizatsiya v Rossii” (Dedollarization in Russia), Vremya (August 7).
- Sprenkle, Case M., 1993, “The Case of the Missing Currency,” Journal of Economic Perspectives, Vol. 7, No. 4, pp. 175–84.

### Inflation dynamics and modeling
- Budina, Nina, Maliszewski, Wojciech, De Menil, Georges, and Turlea, Geomina, “Money, Inflation and Output in Romania, 1992–2000,” Journal of International Money and Finance, forthcoming.
- Celasun, Oya, and Mangal Goswami, 2002, “An Analysis of Money Demand and Inflation in the Islamic Republic of Iran,” IMF Working Paper 03/126 (Washington: International Monetary Fund).
- De Brouwer, Gordon, and Neil. R. Ericsson, 1998, “Modeling Inflation in Australia,” Journal of Business and Economic Statistics, Vol. 16, No. 4, pp. 433–49.
- Kamin, Steven B., and Neil R. Ericsson, 2003, “Dollarization in Post-Hyperinflationary Argentina,” Journal of International Money and Finance, Vol. 22, No. 2, pp. 185–211.
- Kuijs, Louis, 2002, “Monetary Policy Transmission Mechanisms and Inflation in the Slovak Republic,” IMF Working Paper 02/80 (Washington: International Monetary Fund).
- Lissovolik, Bogdan, 2003, “Determinants of Inflation in a Transition Economy: The Case of Ukraine,” IMF Working Paper 03/126 (Washington: International Monetary Fund).
- Maliszewski, Wojciech, 2003, “Modeling Inflation in Georgia”, IMF Working Paper 03/212 (Washington: International Monetary Fund).
- Sacerdoti, Emilio, and Yuan Xiao, 2001, “Inflation Dynamics in Madagascar,” IMF Working Paper 01/168 (Washington: International Monetary Fund).
- Sekine, Toshitaka, 2001, “Modeling and Forecasting Inflation in Japan,” IMF Working Paper 01/82 (Washington: International Monetary Fund).
- Stavrev, Emil, 2003, “The Pass-Through from the Nominal Exchange Rate to Inflation,” in Russian Federation: Selected Issues and Statistical Appendix, IMF Staff Country Report No. 03/142 (Washington: International Monetary Fund), pp. 17–28.
- Taylor, M., 1991, “The Hyperinflation Model of Money Demand Revisited,” Journal of Money, Credit and Banking, Vol. 23, pp. 327–51.
- Williams, Oral, and Olumuyiwa S. Adedeji, 2004, “Inflation Dynamics in the Dominican Republic,” IMF Working Paper 04/29 (Washington: International Monetary Fund).

### Country-specific monetary and financial studies
- Buch, Claudia M., 1998, “Russian Monetary Policy—Assessing the Track Record,” Economic Systems Vol. 22, No. 2, pp. 105–145.
- Buchs, Thierry D., 1999, “Financial Crisis in the Russian Federation. Are Russians Learning to Tango?” Economics of Transition, Vol. 7, No. 3, pp. 687–715.
- Desai, Padma, 2000, “Why Did the Ruble Collapse in August 1998?” American Economic Review, Papers and Proceedings, Vol. 90 (May), pp. 48–52.
- Kharas, Homi, Brian Pinto, and Sergei Ulatov, 2001, “An Analysis of Russia’s 1998 Meltdown: Fundamentals and Market Signals,” Brookings Papers on Economic Activity: 1, pp. 1–50.
- Nikolaenko, Sergey, 1998, “Household Savings” (in Russian), Ekonomicheskiy Zhurnal Vysshoj Shkoloj Ekonomiki, Vol. 2, No. 4, 1998, pp. 5000-5007.
- Nikolić, Milan, 2000, “Money Growth-Inflation Relationship in Postcommunist Russia,” Journal of Comparative Economics, Vol. 28, pp. 108–33.
- Pesonen, Hanna, and Iikka Korhonen, 1999, “The Short and Variable Lags of Russian Monetary Policy,” Russian and East European Finance and Trade, Vol. 35, No. 2, pp. 59–72.
- Rimashevskaya, N.M., 1998, “Household Savings and Domestic Sources of Economic Growth in Russia: Results of Experimental Research” (in Russian), Ekonomika i Matematicheskie Metody, Vol. 24, No. 3.
- Oomes, Nienke, 2003, “Network Externalities and Dollarization Hysteresis: The Case of Russia,” IMF Working Paper 03/96 (Washington: International Monetary Fund).

### Historical and theoretical foundations
- Duesenberry, J., 1950, “The Mechanics of Inflation,” Review of Economics and Statistics, Vol. 32, pp. 144–49.
- Friedman, Milton, and Anna Jacobson Schwartz, 1963, A Monetary History of the United States, 1867–1960 (Princeton, New Jersey: Princeton University Press).
- Hendry, David F., 2003, “Unpredictability and the Foundations of Economic Forecasting,” Working Paper (Oxford, United Kingdom: Oxford University, Economics Department).

*Source: _wp05144 - References*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05144.pdf_
