## 12. Actual and Equilibrium Real Effective Exchange Rate, 1970–2005

## Source details

**Canonical URL:** [12. Actual and Equilibrium Real Effective Exchange Rate, 1970–2005](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06221.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06221.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06221.pdf.json)

---

### Introduction and methodological overview
- The Brazilian real (BRL) appreciated by approximately 40 percent in real effective terms between January 2003 and December 2005.
- The appreciation coincided with a significant improvement in external accounts, driven by a surge in exports and record trade surpluses.
- From a saving–investment perspective, the external adjustment and current account surpluses registered in 2003–05 reflected a strong increase in the national saving rate, notably that of the private sector.
- Methodology:
  - Apply a Behavioral Equilibrium Exchange Rate (BEER) approach.
  - Estimate a vector error-correction (VEC) model for the real effective exchange rate (REER) and potential determinants.
  - Use cointegrating vector(s) to infer the equilibrium path of the REER.
- Main conclusion preview: BEER estimates confirm that key economic fundamentals explain most of the BRL appreciation over 2003–05; this appreciation can be regarded as an “equilibrium phenomenon.”

### The external adjustment (empirical description)
- Current account and net foreign assets:
  - The current account shifted from a deficit of about 4¼ percent of GDP in 1998 to a surplus of 1¾ percent of GDP in 2005.
- Trade and exports:
  - Exports and imports reached record levels while trade and current account surpluses materialized.
  - Twelve-month growth of export volumes rose sharply from the second half of 1999 and remained positive thereafter, though highly volatile.
  - Export market shares recovered slightly during 1999–2001 but began to decline since 2005.
  - A second spurt of export growth began in the latter half of 2002 after a sharp REER depreciation; double-digit export volume growth rates were observed concomitantly with a steady appreciation of the REER during 2003–05.
- Productive capacity and savings:
  - Installed productive capacity in manufacturing expanded following the adoption of the floating exchange rate regime.
  - Gross national savings averaged 21½ percent during 2003–05, compared to about 17 percent in 1998.
  - Gross private savings averaged almost 25 percent of GDP in 2003–05, compared to less than 22 percent in 1998.
  - Public saving improved by about 1½ percentage point of GDP in the same comparison.
- Debt and risk indicators:
  - Public and external debt ratios declined significantly; short-term debt to reserves declined.
  - Composition of domestic public debt improved with elimination of all dollar-indexed debt.
  - Country risk fell, as reflected in EMBI indicators.
- Policy actions and caveats:
  - Reduced debt burden, improved financing conditions, higher export prices and terms of trade, higher national saving rate, and expanded manufacturing capacity imply external equilibrium can be sustained with a more appreciated exchange rate.
  - Caveats: rising manufacturing unit labor costs relative to trading partners, sharp deceleration of export volume growth, and declining export market shares since 2005 raised concerns about further appreciation.
  - The Brazilian Central Bank intervened with large foreign exchange purchases in 2005 and 2006 to strengthen reserves and avoid further appreciation.

### BEER analysis — variables, data, and stationarity
- Endogenous variables used to explain REER:
  - NTT: relative price of nontradables to tradables.
  - TOT: terms of trade.
  - RINTDIFF: real interest rate differentials.
  - NFA: net foreign assets position (ratio to GDP).
  - RELDEBT: relative stock of public domestic debt (general government net debt to GDP in Brazil divided by weighted average for partners).
- Theoretical expectations summarized:
  - NTT: faster productivity in tradables → lower tradable prices → REER appreciation.
  - TOT: deterioration → REER depreciation.
  - RINTDIFF: increase → REER appreciation.
  - NFA: worse NFA → weaker equilibrium currency.
  - RELDEBT: higher relative public debt → equilibrium depreciation via risk premium.
- Data and sample:
  - Annual data, 1970–2004. Data sources include IMF (REER), IMF WEO and IFS, IPEA, and Lane and Milesi-Ferretti (2006) for NFA.
- Stationarity (Augmented Dickey–Fuller tests; lag length by Schwarz, max 8 lags):
  - REER: Level -1.82; First Difference -4.80 **.
  - NTT: Level -1.31; First Difference -5.57 **.
  - TOT: Level -2.50; First Difference -4.96 **.
  - RINTDIFF: Level -2.93*; First Difference -5.72 **.
  - NFA: Level -1.47; First Difference -4.69 **.
  - RELDEBT: Level -1.87; First Difference -5.20 **.
  - ** denotes significance at the 5-percent significance level; * denotes significance at the 10-percent significance level.
  - RINTDIFF showed a p-value below 0.10 at level (borderline), but was treated as I(1) in most specifications.

### Empirical modeling, cointegration, and adjustment
- Method: Vector Error-Correction (VEC) / Johansen cointegration approach with Yt′ = [REERt, NTTt, TOTt, RINTDIFFt, NFAt, RELDEBTt].
- Cointegration results:
  - Johansen tests support one cointegrating relation in main specifications.
  - Adjustment (speed-of-reversion) coefficients (α) estimated in different models:
    - Model I: α = -0.46 [-1.9]
    - Model II: α = -0.54 [-2.1]
    - Model III: α = -0.70 [-4.8]
    - Model IV: α = -0.67 [-4.0]
  - Interpretation: Real exchange rate deviations from equilibrium are corrected relatively fast.
- Selected long-run cointegrating coefficients (REER normalized to 1; t-ratios in brackets):
  - Model I:
    - TOT: -1.49 [-9.0]
    - NTT: -0.46 [-5.9]
    - RINTDIFF: -0.23 [-9.9]
    - RELDEBT: 0.39 [11.3]
    - NFA: -0.78 [-6.8]
  - Model II:
    - TOT: -1.41 [-8.2]
    - NTT: -0.44 [-5.7]
    - RINTDIFF: -0.22 [-9.4]
    - RELDEBT: 0.35 [10.1]
    - NFA: -0.79 [-6.5]
  - Model III (RINTDIFF excluded from cointegrating relation):
    - TOT: -1.73 [-9.3]
    - NTT: -0.39 [-4.6]
    - RELDEBT: 0.38 [9.8]
    - NFA: -1.09 [-8.7]
  - Model IV (RINTDIFF excluded; VARINFL and het. dummy included):
    - TOT: -1.91 [-8.8]
    - NTT: -0.45 [-4.4]
    - RELDEBT: 0.35 [8.2]
    - NFA: -1.20 [-8.7]
- Model diagnostics:
  - Residuals tested for autocorrelation, heteroskedasticity, and normality with satisfactory results in main specifications.

### BEER appraisal of 2003–2005 appreciation — model estimates
- BEER implied appreciation during 2003→2005 across models:
  - Model I: 23%
  - Model II: 23%
  - Model III: 29%
  - Model IV: 31%
- Interpretation:
  - Models I and II imply a BEER appreciation of about 23 percent during 2003→2005.
  - Point estimates indicate BEER was below actual REER in 2005 but within reasonable confidence intervals — supporting that most of the 25 percent REER appreciation observed between 2003 and 2005 was an equilibrium response to improved fundamentals.
  - Model III equilibrium appreciation: about 29 percent during 2003–05; its point estimate implies the real was broadly aligned with fundamentals in 2005.
  - Model IV equilibrium appreciation: 31 percent; this model yields an estimated BEER that corresponds exactly to the observed real exchange rate in 2005.

### Relative contributions of fundamentals to BEER appreciation (2003–05)
- Broad ordering of importance is consistent across specifications.
- Main contributions and magnitudes:
  - Strengthening of net foreign assets (NFA):
    - Ratio to GDP estimated to have risen by about 15 percentage points.
    - Contributed about 60 percent of the estimated movement in the BEER in the period.
  - Improvement in terms of trade (TOT):
    - Combined with high estimated elasticity, generated about 40 percent of the estimated BEER appreciation.
  - Increase in the relative ratio of tradable to nontradable prices (NTT) exerted downward pressure on the equilibrium exchange rate (reflecting a decline in Brazil’s relative productivity of tradables).
  - Interest rate differential (RINTDIFF):
    - Contribution relatively small, reflecting a small absolute value of its coefficient.
    - Interpretation: appreciation owes less to interest differentials per se and more to a decline in investment risk (captured by RELDEBT and other risk measures).
  - Relative public debt ratio (RELDEBT):
    - Fall in RELDEBT accounted for up to 10 percent (on average) of the estimated BEER appreciation between 2003 and 2005.
    - RELDEBT likely understates the full risk-reduction effect because improved debt composition (e.g., virtual elimination of dollar-indexed bonds) also reduced risk.
    - “Brazil risk” measured by the EMBI reached record lows in 2005.
- Table 3 (Relative Contributions of Fundamentals to the BEER Appreciation, 2003–05, in percent of total BEER appreciation) — values preserved as presented:
  - Model I: TOT 23% | NTT 42 | RINTDIFF -15 | RELDEBT 6 | NFA 10
  - Model II: TOT 23% | NTT 41 | RINTDIFF -14 | RELDEBT 5 | NFA 9
  - Model III: TOT 29% | NTT 39 | RINTDIFF -10 | RELDEBT (blank) | NFA 8 | 63
  - Model IV: TOT 25% | NTT 40 | RINTDIFF -11 | RELDEBT (blank) | NFA 7 | 64

### Historical context and additional findings
- 1999 float and peg collapse:
  - The depreciation after the 1999 peg collapse brought the REER very close to its estimated equilibrium level.
  - Equilibrium rate had been driven lower in the late 1990s by a fast-rising domestic public debt to GDP ratio and a deteriorating net foreign assets position.
  - Large capital inflows and rising current account deficits generated about US$110 billion in additional foreign liabilities between 1994–98; servicing these liabilities would require an improvement in the primary current account, implying a need for real exchange rate depreciation.

### Key findings, policy-relevant implications, and risks
- Core findings:
  - Most of the BRL appreciation during 2003–05 can be explained by observable fundamentals and thus can be regarded as largely equilibrium-driven.
  - Important drivers included TOT, NTT (Balassa–Samuelson effects), NFA, RELDEBT, and — where included — RINTDIFF.
  - Speed of adjustment to equilibrium is relatively fast (α estimates range from -0.46 to -0.70).
- Policy-relevant implications:
  - External current account improvement since the 1999 float was supported mainly by higher saving rates, notably private sector saving.
  - Stronger fiscal accounts indirectly supported higher net foreign assets and directly contributed to equilibrium currency strengthening by reducing investment risk.
  - Reduction in investment risk appears to have played a larger role in currency appreciation than real interest rate differentials alone.
  - Central bank interventions (foreign exchange purchases in 2005 and 2006) were used to strengthen reserves and limit further appreciation.
- Risks and caveats:
  - Rising unit labor costs, slowing export volume growth, and declining export market shares since 2005 pose risks to further appreciation and competitiveness.
  - Some model specifications show lower t-ratios (higher uncertainty) when additional exogenous variables (e.g., VARINFL) are included.

*Italic: Content unit: "12. Actual and Equilibrium Real Effective Exchange Rate, 1970–2005" (extracted from the provided PDF chapter).*

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

### _wp06221 - References

### References
- References................................................................................................................................20

### Tables
- 1. Stationarity Tests: ADF Statistics ........................................................................................13
- 2. VEC Models for the Real Effective Exchange Rate ............................................................15
- 3. Relative Contribution of Fundamentals to the BEER Appreciation, 2003–05 ....................18

### Figures
- 1. Brazil: The Real Effective Exchange Rate, Jan 1994–Dec 2005...........................................3
- 2. Brazil: Current Account Balance and Net Foreign Assets, 1994–2005.................................5
- 3. Brazil: Exports, Imports, and the Trade Balance, 1994–2005 ...............................................5
- 4. Brazil: 12-month Growth of Export Volumes and Export Prices 1994–2006.......................6
- 5. Brazil: Export Market Share (index number), 1994:4–2006:1 ..............................................6
- 6. Brazil: Installed Capacity in Manufacturing, (index number), 1994–2005 ...........................7
- 7. Brazil: Saving, Investment, and the Current Account (% GDP), 1994–2005 .......................7
- 8. Brazil: National, Private, and Public Saving (% of GDP), 1994–2005 .................................8
- 9. Brazil: Debt Indicators, 1995–2006 .......................................................................................8
- 10. Brazil EMBI and Overall EMBI, 2002:1–2006:5................................................................9
- 11. Relative Unit Labor Costs, Brazil and Main Partners, 1994–2005....................................10

*Source: _wp06221 - References (PDF).*

### 12. Actual and Equilibrium Real Effective Exchange Rate, 1970–2005.................................16

### 12. Actual and Equilibrium Real Effective Exchange Rate, 1970–2005

### Introduction
- The Brazilian real (BRL) appreciated by approximately 40 percent in real effective terms between January 2003 and December 2005.
- The appreciation coincided with a significant improvement in external accounts, driven by a surge in exports and record trade surpluses.
- From a saving–investment perspective, the external adjustment and current account surpluses registered in 2003–05 reflected a strong increase in the national saving rate, notably that of the private sector.
- The paper applies a Behavioral Equilibrium Exchange Rate (BEER) methodology:
  - Estimate a vector error-correction model for the real effective exchange rate (REER) and potential determinants (fundamentals).
  - Use cointegrating vector(s) to infer the equilibrium path of the REER.
- Main conclusion preview: BEER estimates for Brazil confirm that key economic fundamentals explain most of the BRL appreciation over 2003–05; this appreciation can be regarded as an “equilibrium phenomenon.”

*Source content: introductory discussion and methodological overview.*

### The External Adjustment
- Current account and net foreign assets:
  - The current account shifted from a deficit of about 4¼ percent of GDP in 1998 to a surplus of 1¾ percent of GDP in 2005.
- Trade dynamics:
  - Exports and imports reached record levels while trade and current account surpluses materialized.
  - Twelve-month growth of export volumes rose sharply from the second half of 1999 and remained positive thereafter, though highly volatile.
  - Export market shares recovered slightly during 1999–2001 but began to decline since 2005.
  - A second spurt of export growth began in the latter half of 2002 after a sharp REER depreciation; double-digit export volume growth rates were observed concomitantly with a steady appreciation of the REER during 2003–05.
- Productive capacity:
  - Installed productive capacity in manufacturing expanded following the adoption of the floating exchange rate regime.
- Saving–investment perspective (percentages of GDP):
  - Gross national savings averaged 21½ percent during 2003–05, compared to about 17 percent in 1998.
  - Gross private savings averaged almost 25 percent of GDP in 2003–05, compared to less than 22 percent in 1998.
  - Public saving improved by about 1½ percentage point of GDP in the same comparison.
- Debt and risk indicators:
  - Public and external debt ratios declined significantly; short-term debt to reserves declined.
  - Composition of domestic public debt improved with elimination of all dollar-indexed debt.
  - Country risk fell, as reflected in EMBI indicators.
- Implications:
  - Reduced debt burden, improved financing conditions, higher export prices and terms of trade, higher national saving rate, and expanded manufacturing capacity imply that external equilibrium can be sustained with a more appreciated exchange rate — an equilibrium appreciation.
  - Caveats: rising manufacturing unit labor costs relative to trading partners, sharp deceleration of export volume growth, and declining export market shares since 2005 raised concerns about further appreciation.
  - Brazilian Central Bank intervened with large foreign exchange purchases in 2005 and 2006 to strengthen reserves and avoid further appreciation.

*Source content: empirical description of Brazil’s external adjustment and associated macroeconomic indicators.*

### The Equilibrium Exchange Rate in Brazil (BEER analysis)

A. Variables: motivation and definition
- Endogenous variables used to explain REER:
  - NTT: relative price of nontradables to tradables — measured as (CPI/PPI in Brazil) divided by a similar trade-weighted ratio for Brazil’s main trading partners. Captures Balassa–Samuelson productivity effects and changes in consumer preferences.
  - TOT: terms of trade — ratio of export to import prices in Brazil divided by a similar trade-weighted measure for partners.
  - RINTDIFF: real interest rate differentials — Brazil’s real short-term interest rate minus trade-weighted average of partners’ real short-term rates.
  - NFA: net foreign assets position — ratio to GDP.
  - RELDEBT: relative stock of public domestic debt — general government net debt to GDP in Brazil divided by weighted average for partners.
- Theoretical expectations:
  - NTT: faster productivity in tradables → lower tradable prices → REER appreciation.
  - TOT: deterioration → REER depreciation.
  - RINTDIFF: increase → REER appreciation (uncovered interest parity influence).
  - NFA: higher foreign borrowing (worse NFA) → weaker equilibrium currency.
  - RELDEBT: higher relative public debt → equilibrium depreciation via risk premium.

B. Data and stationarity
- Sample: annual data, 1970–2004.
- Data sources: IMF (REER), IMF WEO and IFS, Brazilian Institute of Applied Economics (IPEA), and Lane and Milesi-Ferretti (2006) for NFA.
- Augmented Dickey–Fuller unit root tests (lag length selected by Schwarz information criteria, max 8 lags). All series appear I(1) (nonstationary in levels, stationary in first difference). Test statistics (ADF):
  - REER: Level -1.82; First Difference -4.80 **.
  - NTT: Level -1.31; First Difference -5.57 **.
  - TOT: Level -2.50; First Difference -4.96 **.
  - RINTDIFF: Level -2.93*; First Difference -5.72 **.
  - NFA: Level -1.47; First Difference -4.69 **.
  - RELDEBT: Level -1.87; First Difference -5.20 **.
  - ** denotes significance at the 5-percent significance level; * denotes significance at the 10-percent significance level.
- RINTDIFF showed a p-value below 0.10 at level (borderline), but was treated as I(1) in most specifications.

C. Empirical modeling and key results
- Method: Vector Error-Correction (VEC) / Johansen cointegration approach with Yt′ = [REERt, NTTt, TOTt, RINTDIFFt, NFAt, RELDEBTt].
- Model selection followed standard procedures; residuals tested for autocorrelation, heteroskedasticity, and normality with satisfactory results.
- Cointegration:
  - Adjusted Johansen tests (trace and max-eigenvalue) support one cointegrating relation in main specifications.
  - Adjustment (speed-of-reversion) coefficients (α) estimated:
    - Model I: α = -0.46 [-1.9]
    - Model II: α = -0.54 [-2.1]
    - Model III: α = -0.70 [-4.8]
    - Model IV: α = -0.67 [-4.0]
  - Interpretation: Real exchange rate deviations from equilibrium are corrected relatively fast.
- Long-run cointegrating coefficients (selected values from VEC models; REER normalized to 1):
  - Model I coefficients (t-ratios in brackets):
    - TOT: -1.49 [-9.0]
    - NTT: -0.46 [-5.9]
    - RINTDIFF: -0.23 [-9.9]
    - RELDEBT: 0.39 [11.3]
    - NFA: -0.78 [-6.8]
  - Model II coefficients:
    - TOT: -1.41 [-8.2]
    - NTT: -0.44 [-5.7]
    - RINTDIFF: -0.22 [-9.4]
    - RELDEBT: 0.35 [10.1]
    - NFA: -0.79 [-6.5]
  - Model III coefficients:
    - TOT: -1.73 [-9.3]
    - NTT: -0.39 [-4.6]
    - RELDEBT: 0.38 [9.8]
    - NFA: -1.09 [-8.7]
    - (RINTDIFF excluded from cointegrating relation in III)
  - Model IV coefficients:
    - TOT: -1.91 [-8.8]
    - NTT: -0.45 [-4.4]
    - RELDEBT: 0.35 [8.2]
    - NFA: -1.20 [-8.7]
    - (RINTDIFF excluded; additional exogenous Varinfl and het. dummy)
- BEER appraisal of 2003–2005 appreciation:
  - Models I and II imply a BEER appreciation of about 23 percent during 2003→2005.
  - Table ∆BEER 2003 →2005 values across models: 23%, 23%, 29%, 31%.
  - Point estimates indicate BEER was below actual REER in 2005 but within reasonable confidence intervals — supporting that most of the 25 percent REER appreciation observed between 2003 and 2005 was an equilibrium response to improved fundamentals.
- Model performance:
  - Models I and II track Brazil’s REER movements well over 1970–2005 (noting measurement distortions around 1990 and 1994 due to hyperinflation and stabilization episodes).
  - The estimated cointegrating vectors help identify the relative importance of terms of trade, productivity (NTT), debt indicators (RELDEBT, NFA), and interest differentials (RINTDIFF) in driving REER dynamics.

*Source content: empirical BEER estimation results, cointegration tests, coefficient estimates, and interpretation.*

### Key findings and implications
- Most of the BRL appreciation during 2003–05 can be explained by observable fundamentals and thus can be regarded as largely equilibrium-driven.
- Important drivers of the equilibrium REER included:
  - Terms of trade (TOT) — significant negative long-run coefficient.
  - Relative prices of nontradables to tradables (NTT) — capturing Balassa–Samuelson type effects.
  - Net foreign asset position (NFA) — negative coefficient implying worse NFA → weaker equilibrium currency.
  - Relative public debt (RELDEBT) — positive coefficient implying higher relative debt → depreciation via risk premium.
  - Real interest differential (RINTDIFF) — contributed to appreciation in specifications where it enters the cointegrating relation.
- Speed of adjustment to equilibrium is relatively fast (α estimates range from -0.46 to -0.70 across specifications).
- Despite equilibrium justification for appreciation, risks remain from rising unit labor costs, slowing export volume growth, and declining export market shares since 2005.

*Source content: synthesis of econometric and descriptive evidence and policy-relevant implications.*

_Italic: Content unit: "12. Actual and Equilibrium Real Effective Exchange Rate, 1970–2005" (extracted from the provided PDF chapter)._

### conclusion that the exchange rate was broadly in line with fundamentals in 2005.

### _wp06221 - conclusion that the exchange rate was broadly in line with fundamentals in 2005

### Model specifications and estimation results
- Model III:
  - Treats the interest differential (RINTDIFF) as an exogenous variable, removing it from the cointegrating vector and including it only among short-term coefficients.
  - Trace and maximum eigenvalue statistics indicate a single cointegrating relation among the five remaining endogenous variables.
  - Estimated coefficients on terms of trade and net foreign assets show a stronger positive impact on the real effective exchange rate (REER).
  - The coefficient on the error-correction term is higher in absolute terms and more significant.
  - The short-run coefficient on the interest rate differential has the expected positive sign and is statistically significant.
  - The estimated equilibrium appreciation implied by the model: the equilibrium in Brazil appreciated by about 29 percent during 2003–05.
  - The point estimate of the BEER obtained through this model implies that the real was broadly aligned with fundamentals in 2005.

- Model IV:
  - Keeps the interest rate differential among exogenous variables and replaces the heterodox plan dummy with VARINFL (variance of monthly inflation each year).
  - VARINFL is stationary according to unit root tests and is continuous, tracing stabilization plans via sudden reductions in monthly inflation rates.
  - t-ratios on the coefficients in the cointegrating vector are lower than in previous specifications, increasing uncertainty around estimates.
  - The equilibrium appreciation implied by this model is 31 percent.
  - The model yields an estimated BEER that corresponds exactly to the observed real exchange rate in 2005.

### Relative importance of fundamentals for 2003–05 REER appreciation
- Order of importance of fundamentals is consistent across model specifications.
- Main contributing factors to BEER appreciation 2003–05:
  - Strengthening of net foreign assets (NFA):
    - Ratio to GDP estimated to have risen by about 15 percentage points.
    - Contributed about 60 percent of the estimated movement in the BEER in the period.
  - Improvement in terms of trade (TOT):
    - Combined with high estimated elasticity, generated about 40 percent of the estimated BEER appreciation.
  - Increase in the relative ratio of tradable to nontradable prices (NTT) exerted downward pressure on the equilibrium exchange rate, possibly reflecting a decline in Brazil’s relative productivity of tradables.
  - Interest rate differential (RINTDIFF) contribution was relatively small, reflecting a small absolute value of its coefficient in the cointegrating equation.
    - Model interpretation: recent impact of interest rate differentials should be analyzed with the risk factor captured by the relative public debt ratio (RELDEBT); the strengthening of the real in 2003–05 owes less to interest differentials per se and more to a decline in investment risk.
  - Relative public debt ratio (RELDEBT):
    - Fall in RELDEBT accounted for up to 10 percent (on average) of the estimated BEER appreciation between 2003 and 2005.
    - RELDEBT introduced to proxy investment risk; it likely underestimates the equilibrium appreciation because risk also declined due to improved debt composition (e.g., virtual elimination of dollar-indexed bonds).
    - “Brazil risk” measured by the EMBI reached record lows in 2005.

- Table 3: Relative Contributions of Fundamentals to the BEER Appreciation, 2003–05 (in percent of total BEER appreciation)
  - Model I: TOT 23% | NTT 42 | RINTDIFF -15 | RELDEBT 6 | NFA 10 | (table ordering preserved as in source)
  - Model II: TOT 23% | NTT 41 | RINTDIFF -14 | RELDEBT 5 | NFA 9
  - Model III: TOT 29% | NTT 39 | RINTDIFF -10 | RELDEBT (blank) | NFA 8 | 63
  - Model IV: TOT 25% | NTT 40 | RINTDIFF -11 | RELDEBT (blank) | NFA 7 | 64

  (Table as presented in source; numerical values preserved exactly as reported.)

### Historical implications and additional findings
- 1999 float and peg collapse:
  - The depreciation after the 1999 peg collapse brought the REER very close to its estimated equilibrium level.
  - Equilibrium rate had been driven lower in the late 1990s by a fast-rising domestic public debt to GDP ratio and a deteriorating net foreign assets position, placing BRL under pressure.
  - Large capital inflows and rising current account deficits generated about US$110 billion in additional foreign liabilities between 1994–98; servicing these liabilities would require an improvement in the primary current account, implying a need for real exchange rate depreciation.

### Concluding remarks and policy-relevant implications
- External current account improvement since the 1999 float:
  - Trade performance benefited from initial REER depreciation, resumption of manufacturing investment, acceleration of global activity, and significant gains in export prices.
  - External adjustment supported mainly by higher saving rates, notably private sector saving.
  - Record trade and current account surpluses coexisted with sharp REER appreciation during 2003–05, suggesting most of the appreciation was an equilibrium response to improved domestic and external conditions.

- Fiscal discipline and risk:
  - Stronger fiscal accounts indirectly supported higher net foreign assets and directly contributed to equilibrium currency strengthening by reducing investment risk.
  - Reduction in investment risk appears to have played a larger role in currency appreciation than real interest rate differentials alone.

- Overall assessment:
  - Model estimates and preliminary 2005 data suggest most of the BRL appreciation during 2003–05 was driven by fundamentals, notably NFA strengthening and more favorable TOT.
  - Models’ point estimates for the equilibrium real exchange rate index fall within reasonable confidence intervals, supporting the conclusion that the BRL was broadly aligned with fundamentals in 2005.

*Source: _wp06221 - conclusion that the exchange rate was broadly in line with fundamentals in 2005.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2006/_wp06221.pdf_