## _wp05202

## Source details

**Canonical URL:** [_wp05202](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05202.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05202.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05202.pdf.json)

---

### Summary and key result
- Studies of China’s real “equilibrium” exchange rate produce a very wide range of estimates, with estimated undervaluation ranging from little or no undervaluation to an undervaluation of nearly 50 percent.
- The paper attributes the wide variation in estimates to:
  - different methodologies used;
  - different explanatory variables included;
  - subjective judgments by researchers; and
  - instability in underlying economic relationships in a rapidly developing economy like China.

### Major empirical approaches
- Two broad approaches dominate the literature:
  - Macroeconomic balance approach:
    - Derives the change in the real exchange rate needed to bring about balance of payments equilibrium.
    - Typical steps: (i) estimate an “underlying” current account position; (ii) establish a “norm” for the external balance (e.g., normal net capital inflows, stable NFA/GDP, or structural saving-investment balance); (iii) compute the gap and convert it into a required real exchange rate change using a trade model and assumed price elasticities.
    - Requires considerable data and subjective judgments; relies on stability of structural relationships.
    - Can be forward-looking but may overstate required exchange rate adjustment because it assumes the entire gap is closed by the real exchange rate change alone.
  - Extended purchasing power parity (PPP) approach:
    - Directly estimates an equation for the level of the real exchange rate using “fundamentals.”
    - Treats the equilibrium real exchange rate as the value predicted by the estimated equation; misalignment is the deviation of actual from predicted.
    - Common fundamentals include the Balassa-Samuelson effect, net foreign assets, terms of trade, and openness.
    - Requires less information but is essentially backward-looking and sensitive to the choice and measurement of explanatory variables.

### Findings using the macroeconomic balance approach (Table 2)
- Results vary considerably depending on how each step is implemented (estimation of underlying current account, choice of norm, trade elasticities, and treatment of processing trade).
- Exact figures reported in selected studies:
  - Goldstein (2004)
    - Underlying Current Account: 2.5 percent of GDP
    - Balance of Payments Norm: -1.5 percent of GDP (based on "normal" capital inflows)
    - Estimated Undervaluation: 15-30 percent
  - Wang (2004) — specification 1
    - Underlying Current Account: 2.1 percent of GDP (average 2000-02)
    - Balance of Payments Norm: 3.1 percent of GDP (savings-investment balance from panel data estimates)
    - Estimated Undervaluation: Small overvaluation
  - Wang (2004) — specification 2
    - Underlying Current Account: 2.1 percent of GDP (average 2000-02)
    - Balance of Payments Norm: 0.98 percent of GDP (stabilizes NFA/GDP at 2001 level)
    - Estimated Undervaluation: Small undervaluation
  - Coudert and Couharde (2005)
    - Underlying Current Account: 2.5 percent of GDP (model based)
    - Balance of Payments Norm: -1.5 percent of GDP (savings-investment balance from panel data estimates)
    - Estimated Undervaluation: 23 percent
- Methodological notes affecting these estimates:
  - Choice of underlying current account (projected 2004 adjusted for overheating and lags; average 2000–2002; potential GDP adjustments and lag structures).
  - Choice of norm (average net capital inflows for 1999–2002; saving-investment balances from different panel estimates; stabilizing NFA/GDP at 2001 level).
  - Choice of trade price elasticities (Goldstein uses elasticities between 0.5 and 1 and explicitly allows for import content of exports at 35–40 percent; Coudert and Couharde do not explicitly allow for processing trade).

### Findings using the extended PPP approach
- The Balassa-Samuelson effect commonly used to explain deviations from PPP: faster productivity growth in traded vs. nontraded sectors causes nontraded relative prices to rise, leading to real appreciation.
- Other commonly included variables: net foreign asset position, terms of trade, openness, and proxies for imperfect substitutability.
- Selected study findings and influences:
  - Frankel (2004) and Coudert and Couharde (2005) interpret a weak Balassa-Samuelson effect for China as evidence of possible undervaluation.
    - Frankel (2004) uses a large cross-section (including developed and developing countries) and finds an estimated undervaluation as large as 36 percent (assessment year 2000 reported in Table 1).
    - Coudert and Couharde (2005) use a smaller group restricted to emerging market economies and report estimates in the range reported in Table 1 (e.g., 18-49 percent for 2002 in the PPP column of Table 1 and 23 percent for 2003 in Table 2).
  - Lee et al. (2005) estimate an extended PPP equation using panel data for 39 countries and include net foreign assets, terms of trade, and a proxy for imperfect substitutability.
  - Wang (2004) estimates an extended PPP equation only for China and finds a significant Balassa-Samuelson effect when including net foreign assets and an openness proxy.
- Differences across studies arise from:
  - Choice of sample (cross-country panels vs. China-only regressions).
  - Choice of proxies for relative productivity (e.g., relative PPP per capita income vs. ratio of consumer prices to producer prices).
  - Definition of the real exchange rate.

### Methodological cautions emphasized in the paper
- All approaches suffer from sensitivity to subjective choices and instability of relationships during rapid structural change.
- The macroeconomic balance approach can compound estimation errors across its three steps and may overstate required exchange rate changes by ignoring general-equilibrium adjustments in output, interest rates, and prices.
- The extended PPP approach is backward-looking and results depend heavily on cross-country “normal” coefficients and variable measurement.

### Estimates of Undervaluation (Extended PPP Approach) — specific figures reported
- PPP per capita income relative to the U.S.: 36 percent — Coudert and Couharde (2005), 145 countries, with population over 1 million, for 2003; Bilateral real exchange rate against the U.S. dollar.
- PPP per capita income relative to the U.S.: 49 percent — Coudert and Couharde (2005), 21 emerging market countries between 1980 Q1 to 2002 Q4; Bilateral real exchange rate against the U.S. dollar.
- Ratio of CPI to PPI: 18 percent — Lee et al. (2005), 39 developed and developing countries, 1980-2003; CPI-based real effective exchange rate.
- Wang (2004), China, 1980-2003: CPI-based real effective exchange rate; Ratio of CPI to PPI; Cumulative current accounts scaled by GDP; Trade openness, dummy variable for 1980-86 — 5 percent.

### Reasons the Extended PPP Approach May Overstate Undervaluation in China
- Balassa-Samuelson assumption on full employment may not apply:
  - The Balassa-Samuelson effect depends on an economy being at full employment or a very high level of employment so rising productivity in traded goods raises wages and spills over to nontraded sector wages.
  - China has a large amount of unemployed and underemployed labor (estimated to be somewhere on the order of 150–200 million workers), and thus does not come close to fulfilling this condition.
- Misspecification of the Balassa-Samuelson proxy:
  - Lack of data leads researchers to proxy the Balassa-Samuelson relative productivity effect by the ratio of China’s consumer price index (CPI) to its producer price index (PPI) relative to the same ratio for China’s trading partners.
  - This proxy presumes a close link between changes in productivity and changes in the CPI/PPI ratio; however, such a link may be less apparent in China because:
    - Elements of China’s CPI, such as utility prices, are still under government control.
    - Housing costs are imputed based on prices in rental markets that are not yet fully developed.
    - There is mismeasurement of price increases, especially in the CPI, because adequate adjustments for improvements in quality particularly of durable goods are not made.
  - Consequently, the link between changes in wages and prices in China may be rather weak.
  - Liberalization of price controls in China may have affected the CPI and the PPI by different magnitudes and at different times, with resulting changes in the CPI/PPI ratio misinterpreted as changes in relative productivity.

### Concluding Remarks on Measuring Equilibrium Exchange Rates
- Given methodological and empirical difficulties, the wide range of estimates for undervaluation of China’s currency, even within a given approach, is not surprising.
- Numerous studies have stoked further debate rather than contributed to forming a consensus, and a more definitive answer is not likely to be forthcoming any time soon.
- Basic limitations in approaches used to estimate an equilibrium exchange rate are formidable:
  - Data problems in the specification of both dependent and explanatory variables remain substantial and will only be solved with time.
  - Key underlying economic relationships determining a country’s exchange rate may stabilize only with time, if at all, making confident estimation difficult.
  - Because of the complexity of these relationships, accurately explaining movements in exchange rates may remain an elusive objective.
- Interim guidance:
  - Estimates of the level of a country’s real exchange rate should continue to be treated with great caution.

*Source: WP/05/202 — Estimating China’s “Equilibrium” Real Exchange Rate, Steven Dunaway and Xiangming Li (October 2005).*

### Section 1

### Estimating China’s “Equilibrium” Real Exchange Rate (Section 1)

### Summary and key result
- Studies of China’s real “equilibrium” exchange rate produce a very wide range of estimates, with estimated undervaluation ranging from little or no undervaluation to an undervaluation of nearly 50 percent.
- The paper attributes the wide variation in estimates to:
  - different methodologies used;
  - different explanatory variables included;
  - subjective judgments by researchers; and
  - instability in underlying economic relationships in a rapidly developing economy like China.

### Major empirical approaches
- Two broad approaches dominate the literature:
  - Macroeconomic balance approach:
    - Derives the change in the real exchange rate needed to bring about balance of payments equilibrium.
    - Typical steps: (i) estimate an “underlying” current account position; (ii) establish a “norm” for the external balance (e.g., normal net capital inflows, stable NFA/GDP, or structural saving-investment balance); (iii) compute the gap and convert it into a required real exchange rate change using a trade model and assumed price elasticities.
    - Requires considerable data and subjective judgments; relies on stability of structural relationships.
    - Can be forward-looking but may overstate required exchange rate adjustment because it assumes the entire gap is closed by the real exchange rate change alone.
  - Extended purchasing power parity (PPP) approach:
    - Directly estimates an equation for the level of the real exchange rate using “fundamentals.”
    - Treats the equilibrium real exchange rate as the value predicted by the estimated equation; misalignment is the deviation of actual from predicted.
    - Common fundamentals include the Balassa-Samuelson effect, net foreign assets, terms of trade, and openness.
    - Requires less information but is essentially backward-looking and sensitive to the choice and measurement of explanatory variables.

### Findings using the macroeconomic balance approach (Table 2)
- Results vary considerably depending on how each step is implemented (estimation of underlying current account, choice of norm, trade elasticities, and treatment of processing trade).
- Exact figures reported in selected studies:
  - Goldstein (2004)
    - Underlying Current Account: 2.5 percent of GDP
    - Balance of Payments Norm: -1.5 percent of GDP (based on "normal" capital inflows)
    - Estimated Undervaluation: 15-30 percent
  - Wang (2004) — specification 1
    - Underlying Current Account: 2.1 percent of GDP (average 2000-02)
    - Balance of Payments Norm: 3.1 percent of GDP (savings-investment balance from panel data estimates)
    - Estimated Undervaluation: Small overvaluation
  - Wang (2004) — specification 2
    - Underlying Current Account: 2.1 percent of GDP (average 2000-02)
    - Balance of Payments Norm: 0.98 percent of GDP (stabilizes NFA/GDP at 2001 level)
    - Estimated Undervaluation: Small undervaluation
  - Coudert and Couharde (2005)
    - Underlying Current Account: 2.5 percent of GDP (model based)
    - Balance of Payments Norm: -1.5 percent of GDP (savings-investment balance from panel data estimates)
    - Estimated Undervaluation: 23 percent
- Methodological notes affecting these estimates:
  - Choice of underlying current account (projected 2004 adjusted for overheating and lags; average 2000–2002; potential GDP adjustments and lag structures).
  - Choice of norm (average net capital inflows for 1999–2002; saving-investment balances from different panel estimates; stabilizing NFA/GDP at 2001 level).
  - Choice of trade price elasticities (Goldstein uses elasticities between 0.5 and 1 and explicitly allows for import content of exports at 35–40 percent; Coudert and Couharde do not explicitly allow for processing trade).

### Findings using the extended PPP approach
- The Balassa-Samuelson effect commonly used to explain deviations from PPP: faster productivity growth in traded vs. nontraded sectors causes nontraded relative prices to rise, leading to real appreciation.
- Other commonly included variables: net foreign asset position, terms of trade, openness, and proxies for imperfect substitutability.
- Selected study findings and influences:
  - Frankel (2004) and Coudert and Couharde (2005) interpret a weak Balassa-Samuelson effect for China as evidence of possible undervaluation.
    - Frankel (2004) uses a large cross-section (including developed and developing countries) and finds an estimated undervaluation as large as 36 percent (assessment year 2000 reported in Table 1).
    - Coudert and Couharde (2005) use a smaller group restricted to emerging market economies and report estimates in the range reported in Table 1 (e.g., 18-49 percent for 2002 in the PPP column of Table 1 and 23 percent for 2003 in Table 2).
  - Lee et al. (2005) estimate an extended PPP equation using panel data for 39 countries and include net foreign assets, terms of trade, and a proxy for imperfect substitutability.
  - Wang (2004) estimates an extended PPP equation only for China and finds a significant Balassa-Samuelson effect when including net foreign assets and an openness proxy.
- Differences across studies arise from:
  - Choice of sample (cross-country panels vs. China-only regressions).
  - Choice of proxies for relative productivity (e.g., relative PPP per capita income vs. ratio of consumer prices to producer prices).
  - Definition of the real exchange rate.

### Methodological cautions emphasized in the paper
- All approaches suffer from sensitivity to subjective choices and instability of relationships during rapid structural change.
- The macroeconomic balance approach can compound estimation errors across its three steps and may overstate required exchange rate changes by ignoring general-equilibrium adjustments in output, interest rates, and prices.
- The extended PPP approach is backward-looking and results depend heavily on cross-country “normal” coefficients and variable measurement.

*Source: WP/05/202 — Estimating China’s “Equilibrium” Real Exchange Rate, Steven Dunaway and Xiangming Li (October 2005).*

### Section 2

### _wp05202 - Section 2

### Estimates of Undervaluation (Extended PPP Approach)
- PPP per capita income relative to the U.S.: 36 percent — Coudert and Couharde (2005), 145 countries, with population over 1 million, for 2003; Bilateral real exchange rate against the U.S. dollar.
- PPP per capita income relative to the U.S.: 49 percent — Coudert and Couharde (2005), 21 emerging market countries between 1980 Q1 to 2002 Q4; Bilateral real exchange rate against the U.S. dollar.
- Ratio of CPI to PPI: 18 percent — Lee et al. (2005), 39 developed and developing countries, 1980-2003; CPI-based real effective exchange rate.
- Wang (2004), China, 1980-2003: CPI-based real effective exchange rate; Ratio of CPI to PPI; Cumulative current accounts scaled by GDP; Trade openness, dummy variable for 1980-86 — 5 percent.

### Reasons the Extended PPP Approach May Overstate Undervaluation in China
- Balassa-Samuelson assumption on full employment may not apply:
  - The Balassa-Samuelson effect depends on an economy being at full employment or a very high level of employment so rising productivity in traded goods raises wages and spills over to nontraded sector wages.
  - China has a large amount of unemployed and underemployed labor (estimated to be somewhere on the order of 150–200 million workers), and thus does not come close to fulfilling this condition.
- Misspecification of the Balassa-Samuelson proxy:
  - Lack of data leads researchers to proxy the Balassa-Samuelson relative productivity effect by the ratio of China’s consumer price index (CPI) to its producer price index (PPI) relative to the same ratio for China’s trading partners.
  - This proxy presumes a close link between changes in productivity and changes in the CPI/PPI ratio; however, such a link may be less apparent in China because:
    - Elements of China’s CPI, such as utility prices, are still under government control.
    - Housing costs are imputed based on prices in rental markets that are not yet fully developed.
    - There is mismeasurement of price increases, especially in the CPI, because adequate adjustments for improvements in quality particularly of durable goods are not made.
  - Consequently, the link between changes in wages and prices in China may be rather weak.
  - Liberalization of price controls in China may have affected the CPI and the PPI by different magnitudes and at different times, with resulting changes in the CPI/PPI ratio misinterpreted as changes in relative productivity.

### Concluding Remarks on Measuring Equilibrium Exchange Rates
- Given methodological and empirical difficulties, the wide range of estimates for undervaluation of China’s currency, even within a given approach, is not surprising.
- Numerous studies have stoked further debate rather than contributed to forming a consensus, and a more definitive answer is not likely to be forthcoming any time soon.
- Basic limitations in approaches used to estimate an equilibrium exchange rate are formidable:
  - Data problems in the specification of both dependent and explanatory variables remain substantial and will only be solved with time.
  - Key underlying economic relationships determining a country’s exchange rate may stabilize only with time, if at all, making confident estimation difficult.
  - Because of the complexity of these relationships, accurately explaining movements in exchange rates may remain an elusive objective.
- Interim guidance:
  - Estimates of the level of a country’s real exchange rate should continue to be treated with great caution.

*Source: _wp05202 - Section 2*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05202.pdf_