## _wp04142

## Source details

**Canonical URL:** [_wp04142](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04142.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04142.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04142.pdf.json)

---

### I. Introduction — scope and context
- Nepal has had a fixed exchange rate regime since 1960 and has sustained the current level of its peg toward the Indian rupee since 1993.
- Only 35–40 percent of Nepal’s GDP consists of traded goods.
- Nepal finalized WTO accession talks in September 2003.
- The Multifiber Agreement (MFA) phase out in 2005 may disadvantage Nepal after the lifting of quota restrictions on Indian and Chinese exports.
- The paper does not analyze the costs and benefits of a fixed-versus-flexible regime for Nepal, nor does it analyze the level of the current peg.
- Caveat: "The export price index for Nepal needed to be constructed. The construction methodology introduced a bias in favor of the U.S. dollar, leading to an overestimation of the influence of the U.S. dollar on export prices."

### II. Vulnerabilities and transmission channels
- Narrow export base and pegged exchange regime increase vulnerability to:
  - Price shocks in traded goods.
  - Price pressures arising from exchange rate fluctuations in major currencies.
- Despite the peg to the Indian rupee, fluctuations in the dollar and the euro affect the Nepalese economy through the traded-goods sector.

### III. Nepal’s traded-goods sector — composition and partners
- Major overseas export commodities (FY 2001/02, overseas exports exclude India; amounts in Nrs millions):
  - Readymade garments: 7,752 — 42.0 percent
  - Woolen carpet: 6,109 — 33.2 percent
  - Woolen and pashmina goods: 1,852 — 10.1 percent
  - Hides and goatskin: 451 — 2.5 percent
  - Other goods: 2,245 — 12.2 percent
  - Total: 18,409 — 100.0 percent
- Major trading partners (FY 2001/02, In millions of Nepalese rupees):
  - Exports:
    - India: 28,865 — 60 percent
    - United States: 9,378 — 19 percent
    - Germany: 4,043 — 8 percent
    - United Kingdom: 809 — 2 percent
    - Italy: 567 — 1 percent
    - Japan: 492 — 1 percent
    - Other countries: 4,142 — 9 percent
    - Total: 48,296 — 100 percent
  - Imports:
    - India: 45,364 — 43 percent
    - Singapore: 7,846 — 7 percent
    - Malaysia: 4,818 — 5 percent
    - Saudi Arabia: 4,573 — 4 percent
    - China: 4,316 — 4 percent
    - Switzerland: 4,206 — 4 percent
    - Other countries: 34,380 — 33 percent
    - Total: 105,503 — 100 percent
- Sectoral export destinations and product links:
  - Garments are exported almost exclusively to the U.S. market.
  - Carpets are exported to Germany.
  - Vegetable ghee is exported to India.
- Note on trade exposure calculation: "Calculated as exports plus imports, excluding re-exports, divided by GDP."

### IV. Methodology and model framing
- The paper applies the Sjaastad (2000) model, in which basket weights are chosen according to the relative price-making power that countries have over the prices of traded goods of a given country.
- Unlike many prior studies that assumed a basket a priori and found trade-weighted baskets optimal, the Sjaastad framework allows corner solutions (a pure U.S. dollar peg or a pure Indian rupee peg) based on price-making power rather than only trade shares.
- Estimation approach and data:
  - Sjaastad equation (3): ∑j Θj (Pj + Ejx) = PTx, with ∑j Θj = 1; variables in natural logarithms.
  - All variables converted to U.S. dollar terms for estimation; monthly exchange rate averages used.
  - Separate estimates for import-price index (IMPF) and export-price index (EXPF) produce import-Θ and export-Θ; overall Θ lies between these estimates.
  - ARDL model used to capture lagged price effects; data monthly; augmented Dickey-Fuller tests indicated variables in levels were I(2) or I(1), so second differencing was used.

### V. Estimation results and interpretation
- Imports (IMPF = PTF):
  - Countries tested included Japan, India, euro area, United Kingdom, United States, and others.
  - Except for India, none of these countries had a significant impact on Nepal’s import-good prices.
  - Restricted regression using United States, India, and euro area could not reject the unit-sum restriction for Θ’s (F = 0.86; significance level = 0.36).
  - Calculated import-side price-making power:
    - India: 106 percent (interpreted as sole price maker for import prices of Nepal in this regression).
  - Additional test: hypothesis that Θ for India equals 1 could not be rejected (F = 0.00; significance level = 0.97).
  - Trade shares context: Nepal imports 43 percent of its goods from India; Singapore accounts for 7 percent of total imports; United States accounts for 2 percent of total imports.
- Exports (EXPF = PTF):
  - Export-side estimation found only the United States and India had significant impacts on Nepalese export prices (euro area dropped due to insignificance).
  - Estimated export price-making powers:
    - United States: 71 percent
    - India: 29 percent
  - Trade shares context: 60 percent of Nepalese exports go to India and 19 percent go to the United States.
  - Caveat: The constructed export index likely biases results in favor of the United States (using U.S. import prices as proxies), thereby overestimating U.S. influence and underestimating euro area and India influence.
- Overall price-making power (using trade-share weighting: imports = 69 percent of traded goods, exports = 31 percent):
  - Θ_USA_Nepal = 31%⋅(exports Θ_USA) + 69%⋅(imports Θ_USA) = 22 percent
  - Empirical summary:
    - Import prices of Nepal are effectively determined solely by the Indian currency area.
    - Export prices of Nepal receive substantial influence from the United States (estimated up to 71 percent under the export-index construction), likely an upper bound.
    - Overall, the United States exerts a non-negligible influence on Nepal’s traded-goods prices — estimated overall price-making power of the United States is 22 percent given current trade composition (69 percent imports, 31 percent exports).

### VI. Policy implications and recommendations
- Current assessment:
  - "The Indian rupee currently appears to be a reasonable anchor for the Nepalese rupee."
- Forward-looking recommendations:
  - As Nepal opens its economy and increases its exports, Nepal may want to reevaluate whether pegging to a currency basket with a positive weight for the U.S. dollar may be a better policy option.
  - Alternatively, Nepal can evaluate the costs and benefits of moving toward a more flexible exchange rate regime.
  - Practical suggestion: consider a currency-basket peg with weights γj chosen equal to estimated jxΘ’s to eliminate external price shocks to inflation and real interest rates.
- Rationale:
  - Pegging only to the Indian rupee currently eliminates shocks on import prices but not on export prices; as exports rise relative to imports, shocks transmitted via export-price channels will gain importance.
  - If Nepal had pegged to a basket including the U.S. dollar, recent depreciations of the U.S. dollar against the Indian rupee would have automatically led to some depreciation of the Nepalese rupee with the U.S. dollar, mitigating adjustment pressures.

### VII. Conclusion
- Nepal is presently appropriately pegged to the Indian rupee given the dominant Indian influence on import prices.
- However, the estimated nontrivial influence of the U.S. dollar on export prices and the potential for export growth imply that a future reevaluation toward a basket that assigns a positive weight to the U.S. dollar may be warranted to better insulate domestic inflation and real rates from external shocks.

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04142.pdf*

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

### _wp04142 - References

### I. Introduction — scope and context
- Nepal has had a fixed exchange rate regime since 1960 and has sustained the current level of its peg toward the Indian rupee since 1993.
- Only 35–40 percent of Nepal’s GDP consists of traded goods.
- Nepal finalized WTO accession talks in September 2003.
- The Multifiber Agreement (MFA) phase out in 2005 may disadvantage Nepal after the lifting of quota restrictions on Indian and Chinese exports.
- The paper does not analyze the costs and benefits of a fixed-versus-flexible regime for Nepal, nor does it analyze the level of the current peg.
- Caveat: "The export price index for Nepal needed to be constructed. The construction methodology introduced a bias in favor of the U.S. dollar, leading to an overestimation of the influence of the U.S. dollar on export prices."

### II. Vulnerabilities and transmission channels
- Narrow export base and pegged exchange regime increase vulnerability to:
  - Price shocks in traded goods.
  - Price pressures arising from exchange rate fluctuations in major currencies.
- Despite the peg to the Indian rupee, fluctuations in the dollar and the euro affect the Nepalese economy through the traded-goods sector.

### III. Nepal’s traded-goods sector — composition and partners
- Major overseas export commodities (FY 2001/02, overseas exports exclude India):
  - Readymade garments: 7,752 (In Nrs millions), 42.0 percent
  - Woolen carpet: 6,109 (In Nrs millions), 33.2 percent
  - Woolen and pashmina goods: 1,852 (In Nrs millions), 10.1 percent
  - Hides and goatskin: 451 (In Nrs millions), 2.5 percent
  - Other goods: 2,245 (In Nrs millions), 12.2 percent
  - Total: 18,409 (In Nrs millions), 100.0 percent
- Major trading partners (FY 2001/02, In millions of Nepalese rupees):
  - Exports:
    - India: 28,865 — 60 percent
    - United States: 9,378 — 19 percent
    - Germany: 4,043 — 8 percent
    - United Kingdom: 809 — 2 percent
    - Italy: 567 — 1 percent
    - Japan: 492 — 1 percent
    - Other countries: 4,142 — 9 percent
    - Total: 48,296 — 100 percent
  - Imports:
    - India: 45,364 — 43 percent
    - Singapore: 7,846 — 7 percent
    - Malaysia: 4,818 — 5 percent
    - Saudi Arabia: 4,573 — 4 percent
    - China: 4,316 — 4 percent
    - Switzerland: 4,206 — 4 percent
    - Other countries: 34,380 — 33 percent
    - Total: 105,503 — 100 percent
- Sectoral export destinations and product links:
  - Garments are exported almost exclusively to the U.S. market.
  - Carpets are exported to Germany.
  - Vegetable ghee is exported to India.
- Note on trade exposure calculation: "Calculated as exports plus imports, excluding re-exports, divided by GDP."

### IV. Estimation results and interpretation
- Import prices in Nepal are solely influenced by the Indian currency area.
- Fluctuations between third currencies (U.S. dollar/rupee, euro/rupee) do not lead to inflationary or deflationary pressures in Nepal’s import prices.
- Nepalese export prices do experience shocks from the U.S. dollar currency area. Specifically:
  - When the U.S. dollar depreciates toward the Indian rupee, pressure builds up on export prices because of the price-making power that the U.S. dollar has over the prices of Nepalese export goods.

### V. Policy implications and recommendations
- Current situation:
  - "The Indian rupee currently appears to be a reasonable anchor for the Nepalese rupee."
- Going forward:
  - As Nepal opens its economy and increases its exports, Nepal may want to reevaluate whether pegging to a currency basket with a positive weight for the U.S. dollar may be a better policy option.
  - Alternatively, Nepal can also evaluate the cost and benefits of moving toward a more flexible exchange rate regime.

### VI. Methodology and model framing
- The paper applies the Sjaastad (2000) model, in which basket weights are chosen according to the relative price-making power that countries have over the prices of traded goods of a given country.
- Unlike many prior studies that assumed a basket a priori and found trade-weighted baskets optimal, the Sjaastad framework allows corner solutions (a pure U.S. dollar peg or a pure Indian rupee peg) based on price-making power rather than only trade shares.

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04142.pdf*

### 23.3 percent of total exports go to Singapore, and 15.4 percent of total imports come from

### _wp04142 - 23.3 percent of total exports go to Singapore, and 15.4 percent of total imports come from

### Background and motivation
- 23.3 percent of total exports go to Singapore, and 15.4 percent of total imports come from Singapore; however, Singapore does not determine world prices of Malaysia’s traded goods such as electrical machinery, crude petroleum, telecommunications, vegetable oil, etc.
- Because Singapore’s price-making power over Malaysia is likely to be low, the Singapore dollar does not need to be present in the Malaysian currency basket, even though optimal-currency-basket literature would assign it a significant weight based on trade shares.
- Recent optimal-basket studies (Ito, Ogawa, and Sasaki (1998)) build micro-founded models where oligopolistic exporters set export prices endogenously; price “stickiness” is a result of optimizing behavior, and the optimal solution minimizes fluctuations in trade balances.

### The Sjaastad model (theoretical framework)
- The paper uses the Sjaastad (1998 and 2000) model to determine currency bloc membership for a small country (example focus: Nepal).
- Key concept: the optimal weights for a currency basket are the relative price-making powers (Θ) that major currency countries have over the small country’s traded-goods prices.
- Equation (1) (in the source) links inflation and real interest rate of small country x pegged to country k to:
  - the anchor-country inflation (Πk),
  - exchange rate changes (Ėkx),
  - and terms involving real exchange rate changes Ṙjk weighted by jxΘ.
- Implication: even under a credible exchange rate rule, equality of real interest rates and inflation is not assured unless external real-exchange-rate shocks are eliminated.
- If jxΘ are large for multiple countries, pegging to a currency basket with weights γj = jxΘ eliminates external price shocks; inflation in the small country becomes the world inflation rate ΠW = ∑M j Πj Θj and the real interest rate becomes the weighted average of major-country real rates.

### Estimation method (practical implementation)
- Starting point: Sjaastad equation (3): ∑j Θj (Pj + Ejx) = PTx, with ∑j Θj = 1; variables are in natural logarithms.
- Converted all variables to U.S. dollar terms (currency i = U.S. dollar) for estimation; monthly exchange rate averages used for conversion.
- Approach: estimate the Θ’s separately using import-price index (IMPF) as proxy for traded-goods prices and export-price index (EXPF) as alternative proxy, yielding import-Θ and export-Θ; overall Θ lies between these estimates.
- Parameterization: ARDL model (equation (6)) to capture lagged price effects; equations (7) and (8) show how estimated coefficients map to jxΘ’s.
- Data: monthly price data; augmented Dickey-Fuller tests indicate variables in levels are either I(2) or I(1), so second differencing was used to achieve stationarity.

### Estimation results — imports (IMPF = PTF)
- Countries tested on RHS variables included Japan, India, euro area, United Kingdom, United States, and others.
- Finding: Except for India, none of these countries had a significant impact on Nepal’s import-good prices.
- Table 3 (ADF tests, sample 01/2000–05/2003) indicates ∆IMPF Nepal and ∆EXPF Nepal: Unit root rejected; some foreign PF series showed unit-root with drift or unit-root rejected as indicated in table.
- Restricted regression using United States, India, and euro area (Table 4) could not reject the unit-sum restriction for Θ’s (F = 0.86; significance level = 0.36).
- Calculated import-side price-making power:
  - India: 106 percent (interpreted as sole price maker for import prices of Nepal in this regression).
- Additional test: the hypothesis that Θ for India equals 1 could not be rejected (F = 0.00; significance level = 0.97), confirming India as sole price maker for Nepal’s import goods.
- Trade shares context: Nepal imports 43 percent of its goods from India; Singapore accounts for 7 percent of total imports; United States accounts for 2 percent of total imports.

### Estimation results — exports (EXPF = PTF)
- Export-side estimation found only the United States and India had significant impacts on Nepalese export prices (euro area dropped due to insignificance).
- Table 5 regression results (sample 01/2000–05/2003) produce:
  - United States price-making power over export goods: 71 percent.
  - India price-making power over export goods: 29 percent.
- Trade shares context: 60 percent of Nepalese exports go to India and 19 percent go to the United States.
- Caveat: The constructed export index likely biases results in favor of the United States (using U.S. import prices as proxies), thereby overestimating U.S. influence and underestimating euro area and India influence.
- Using trade-share weighting (imports = 69 percent of traded goods, exports = 31 percent), overall U.S. price-making power on Nepal’s traded goods is calculated as:
  - Θ_USA_Nepal = 31%⋅(exports Θ_USA) + 69%⋅(imports Θ_USA) = 22 percent
  - Explicit formula in source: %22%69%31=Θ⋅+Θ⋅=Θ ExportsImports USA Nepal

### Key empirical findings
- Import prices of Nepal are effectively determined solely by the Indian currency area (India has full price-making power on imports).
- Export prices of Nepal receive substantial influence from the United States (estimated up to 71 percent under the export-index construction), though this is likely an upper bound due to data/proxy bias.
- Overall, the United States exerts a non-negligible influence on Nepal’s traded-goods prices — estimated overall price-making power of the United States is 22 percent given current trade composition (69 percent imports, 31 percent exports).

### Policy implications and recommendations
- Current policy: Nepal’s peg to the Indian rupee is reasonable given imports are dominated by India and import prices are determined by India.
- Forward-looking recommendation: As Nepal opens its economy and exports grow, Nepal should reevaluate whether pegging to a currency basket that includes a positive weight for the U.S. dollar may become preferable.
- Rationale:
  - If Nepal had pegged to a basket including the U.S. dollar, recent depreciations of the U.S. dollar against the Indian rupee would have automatically led to some depreciation of the Nepalese rupee with the U.S. dollar, mitigating adjustment pressures.
  - Pegging only to the Indian rupee currently eliminates shocks on import prices but not on export prices; as exports rise relative to imports, shocks transmitted via export-price channels will gain importance.
- Practical suggestion: consider a currency-basket peg with weights γj chosen equal to estimated jxΘ’s to eliminate external price shocks to inflation and real interest rates.

### Conclusion
- Nepal is presently appropriately pegged to the Indian rupee given the dominant Indian influence on import prices.
- However, the estimated nontrivial influence of the U.S. dollar on export prices and the potential for export growth imply that a future reevaluation toward a basket that assigns a positive weight to the U.S. dollar may be warranted to better insulate domestic inflation and real rates from external shocks.

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04142.pdf*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2004/_wp04142.pdf_
