Journal Issue
Share
Article

India

Author(s):
International Monetary Fund
Published Date:
February 2008
Share
  • ShareShare
Show Summary Details

I. Competitiveness and Exchange-Rate Policy1

1. This paper presents background information to the staff’s assessment of competitiveness and exchange rate policy in India. It examines the following questions:

  • How has India’s competitiveness evolved in the recent period?

  • How have India’s exports evolved, and what have been the driving factors?

  • How does the CGER framework assess the level of India’s exchange rate?

  • How has exchange rate policy been conducted, and what effects has it had?

While the rupee has appreciated significantly in the recent period, export performance has nonetheless remained favorable, driven in part by structural factors, and the rupee does not appear to be out of line with fundamentals. Intervention has been sizeable, but recent empirical analysis produces no evidence that it has influenced the level or rate of change in the exchange rate (there is modest evidence that it may have dampened volatility, consistent with the authorities’ stated aims).

A. How Has Competitiveness Evolved?

2. The rupee has appreciated significantly during 2007, raising concerns about competitiveness. In nominal bilateral terms vis-a-vis the dollar, the appreciation has been particularly notable, reaching successive nine-year highs as it rose about 12 percent. Although the increase has been less in nominal and real effective terms—only about 7–7½ percent—the appreciation of the real effective rupee has taken it out of the historical range in which it fluctuated during most of this decade.

3. An examination of India’s own historical experience and that of other countries supports the idea that the exchange rate is not the sole determinant of competitiveness or export performance. In particular, a weakening rupee does not guarantee that exports will grow faster, as evidenced by India’s experience during the 1970s and 1980s: despite the fact that the rupee lost more than half its value in real terms against the U.S. dollar, India’s share in world trade fell from 0.75 percent to about 0.5 percent. The international experience supports a similar conclusion. In recent years, Asian exports have been growing rapidly despite significant real currency appreciation. In Korea, for example, the won appreciated by about 23 percent in real effective terms between 2003 and 2006—and export growth averaged 20 percent per year. Exports in Indonesia and Thailand have also grown rapidly despite large appreciation.2

4. India has also managed to expand its exports rapidly despite real appreciation. Since 1992, the rupee has been on a mildly appreciating trend (though exhibiting considerable volatility), and India’s share in world goods exports has roughly doubled. It is interesting to note that the fastest export growth since 1974—30 percent—occurred in 2005, when the rupee appreciated by over 4 percent in real effective terms.

5. These examples show that export performance clearly depends on factors beyond the value of an economy’s currency, with productivity growth particularly important in this regard. The experience of some of Asia’s fastest growing emerging economies is particularly noteworthy. In Korea, for example, productivity growth in industry averaged 6.2 percent per year during 1972–2004, compared to 2.1 percent in the United States and 2.4 percent in Japan.3 China’s productivity grew even faster, averaging 6.8 percent per year.

6. India’s productivity growth is indeed rapid compared with other countries. The September 2006 World Economic Outlook found that India’s total factor productivity growth has averaged about 3⅓ percent in recent years, which within Asia is only exceed by China. Other recent growth accounting exercises have found TFP growth for India in the range of 3.2–3.5 percent for the recent period (see Oura, 2007). Studies suggest that the rapid productivity growth—and more generally, India’s growth takeoff—is the product of reforms launched in the 1990s, in which licensing requirements, financial activities, and FDI were successively deregulated, and trade tariffs slashed. More recently, productivity has likely been spurred by the wave of corporate restructuring that occurred earlier this decade, which brought down corporate leverage to low levels.

7. At the same time, selected industries have lagged in export performance, but these are generally industries where productivity has lagged as well. These industries include gems and jewelery, textiles, and agriculture; notably, total factor productivity growth in agriculture is estimated at only one-half that of industry and one-eighth that in services.4 Moreover, this recent pattern of export performance appears to be an extension of past trends, suggesting that it may not have that much to do with the recent movement in the exchange rate. This chapter next turns to a closer examination of this longer-term phenomenon.

B. Trends in Export Performance

8. Between 1988 and 2005, India’s real exports of goods increased 500 percent, doubling India’s share in world exports and far outstripping GDP growth (Figure I.5). Using disaggregated customs data on India’s trade over this time period, this section describes the pattern of this impressive export performance, including the change in the structure of exports, notably the breadth of products and markets versus specialization, and the shift in the value chain towards more skill-intensive and capital-intensive products. Analysis of these structural changes puts in perspective the relative importance of rupee appreciation in driving export performance.

Figure I.1.India: Effective Exchange Rates

(2000 = 100)

Source: Fund staff estimates.

Figure I.2.India: REER and Share in World Goods Export, 1992-2006

Source: IMF, Information Notice System and Direction of Trade Statistics.

1/ Increase in REER denotes appreciation.

Figure I.3.Industrial Productivity: China, Korea, Japan, and the United States

(1972 = 100)

Source: WEO (April 2007).

Figure I.4.Total Factor Productivity: China, India, Japan, and the United States

(1992 = 100)

Source: WEO (April 2007).

Figure I.5.Export Performance and Real GDP

(1988 = 100)

9. The data source for this analysis is the UN COMTRADE database which provides disaggregated bilateral trade flows for a large number of countries. The data on Indian exports at the HS 6-digit product level (about 5,000 products), is available on an annual basis from 1988 to 2005.5 Products were matched to industry (NIC-1987) codes6 and skill-intensity and capital-intensity at the industry level was obtained from the Annual Survey of Industries.

10. At an aggregate level, India’s export growth picked up sharply in this decade, accompanied by a shift in the composition of India’s exports. Figure I.6 decomposes total merchandise exports by broad industrial groups (SITC-1) at three points in time: 1988, 1999 and 2005. The decline in the relative importance of agricultural products (SITC 0, 4) and light manufacturing (SITC 6) was offset by a shift towards the more sophisticated machinery and capital goods manufacturing (SITC 7) and chemicals (SITC 5). While the exports of raw materials, mineral fuels and oils (SITC 2, 3)were on a steadily declining path for some time, 1999 marked a reversal of this trend as exports of petroleum oils consistently outperformed the average rate of export growth. The change in the export structure in India is not nearly as pronounced as in China where the share in total exports accounted for by machinery and capital goods nearly tripled in 13 years to reach more than 30 percent (see Amiti and Freund, 2007). Yet, the evolution of India’s export composition suggests that India is slowly and steadily moving up the value chain.

Figure I.6.Sectoral Composition of Exports

11. At a more disaggregated level, India continued to expand its already rather diversified product and destination base. The number of goods exported increased gradually from 3,662 in 1998 to 4,687 in 2005 (Figure I.7) out of a total of 5,000 distinct HS-6 product categories. India’s exports reached 225 markets in 2005, compared to 165 in 1988, with countries such as China, Singapore, United Arab Emirates receiving a substantially larger share of total exports relative to the past. The average number of destinations per product tripled to 30 countries (Figure I.8). Following recent convention in the trade literature (Schott, 2004), products exported to different destinations can be treated as different “varieties.”7 The growth in number of products and destinations bolstered a sharp increase in the number of varieties exported, especially in the last five years covered here (Figure I.7).

Figure I.7.Number of Products and Varieties

Figure I.8.Number of Markets per Product

12. But how important, in terms of value, are these new products and varieties? To answer this, one may decompose the growth in exports into growth in existing products/varieties (the intensive margin) and new product/varieties (the extensive margin), as in Hummels and Klenow (2005).

Where V T2000 is the volume of India’ exports in 2005; vi, 2000 is the value of exports of good/variety i in 2005; P1988 is the set of products/varieties India exported in 1988.

The first term in the equation captures the contribution of the intensive margin (i.e., exporting more of the same goods/varieties), while the second term represents the contribution of the extensive margin (the new goods/varieties exported).

  • The extensive margin in terms of products (or the 1000 products India added to its export bundle) accounted for 9 percent of the increase in exports in India between 1988 and 2005. The contribution of new products appears to be relatively small owing to the fact that India already exported three-fourths of the possible product categories by 1988, leaving little room for expansion. As suggested by Amiti and Freund (2007), further downward bias may be introduced by keeping the set of possible products constant over time, while new exports could be in different sub-classifications.

  • In terms of varieties, on the other hand, the extensive margin contributed more than 60 percent of the growth in exports, suggesting that half of the increase in exports came from exporting existing products to new destinations. This finding is in contrast with the experience of China, where new varieties accounted for only 16 percent of export growth over the 1992–2005 period (Amiti and Freund, 2007).

13. Consistent with the introduction of new products and varieties, the concentration of India’s exports declined. Figure I.9 plots the share of the top 5, top 10, top 50, top 100, and top 500 export products in 1988, 1999, and 2005. Over time, a smaller share of total exports is accounted for by a given number of products. This pattern is confirmed by a decline in the Gini coefficient of export distribution, which decreased from 0.95 to 0.91 over this time period.

Figure I.9.Concentration of Exports

(Shares of top products)

14. In addition, exports have become more skill and capital intensive, suggesting a shift towards products higher in the value chain. To measure the change in skill-intensity of India’s exports, HS 6-digit products are matched to industry codes used in the Annual Survey of Industries. Each product is assigned the skill intensity of its corresponding industry, where skill intensity is measured as the ratio of non-production to total workers in the industry as reported in 1991. Similarly, the industry’s capital to labor ratio is used to capture the capital intensity of the product. Following Zhu and Trefler (2005), products are ranked from those that use the least amount of skilled labor/capital to those that use the most. The cumulative export shares are then plotted against the skill/capital intensity in 1988, 1999 and 2005. A shift to the right implies an increase in the skill/capital intensity of India’s exports. Figures I.10 and I.11 present the findings. For both skill and capital intensity, one sees a marked shift of the curve to the right, especially after 1999. India’s exports are becoming both more skilled-labor intensive and more capital-intensive.

Figure I.10.Skill Intensity of India’s Exports

Figure I.11.Capital Intensity of India’s Exports

15. Similarly, the sophistication of India’s exports marked a small improvement. A measure of sophistication of a country’s export basket, suggested by Rodrik (2005), assigns an “income” coefficient to each product based on the weighted average of the incomes of the countries exporting the product. These “income” coefficients of the various products are averaged using the export shares in a particular country’s export basket as weights to arrive at the income level “embedded” in the exports. Figure I.12 plots the log of this measure of export sophistication for India over time, together with the evolutions of India’s per capita real GDP. Even as early as 1988, the level of sophistication of India’s exports was rather high relative to other countries at a similar level of development (Rodrik, 2007). Though outpaced by the growth in the standard of living, the sophistication of India’s export bundle slowly creeps up, particularly in the last 5 years covered in this study.

Figure I.12.Sophistication of India’s Exports

16. The shift in the composition of exports is significant, as it bears on the likely impact of rupee appreciation on competitiveness, and in particular on exporters’ pricing power in global markets. Exporters of more traditional manufactured goods (for example, textiles and leather goods), who produce commoditized goods and thus face stiff international competition, are likely to come under considerable pressure. While service sector exports are also likely to be impacted by the rising rupee, demand for these products is growing rapidly, suggesting that these exports are likely to continue growing too. In addition, some of the newer IT and outsourcing services may be more robust to exchange rate appreciation. In sectors where there is a focus on tailor-made solutions, exporters are likely to command more pricing power, since purchasers of these services are unlikely to switch suppliers on the basis of relatively moderate exchange rate changes.

C. Perspectives on the Level of the Exchange Rate

17. Analysis by the Fund staff’s Consultative Group on Exchange Rates (CGER) analysis finds the exchange rate to be “close to equilibrium.” As the three approaches (macro balance, external sustainability, and reduced form equilibrium RER) all have similar findings, the macro balance approach (MB) is detailed here.8

18. The MB approach asks two questions. First, how far is the underlying current account balance from a benchmark level? Second, how large of an exchange rate adjustment would be required to move the current account balance from the underlying to the benchmark level? Each step is now addressed in turn.

19. The “underlying current account” is the medium-term level of the current account purged of temporary factors. For example, the projected medium-term current account deficit implicitly incorporates the lagged effect of historical exchange rate adjustments. For India, such adjustments result in an underlying deficit estimate of 2¾ percent of GDP.

20. The benchmark current account is calculated based on characteristics of the economy in question. In particular, a larger fiscal deficit translates into a smaller current account balance (lower domestic savings); a large and rising share of prime-age savers implies a larger balance (higher domestic savings); higher oil prices (for a net oil importer like India) imply a lower balance; and rapid growth and a lower per capita income imply a smaller balance, consistent with the idea that developing countries import capital as part of their development trajectories.

21. for India, the fiscal balance, demographics and its level of development are all important drivers of the current account benchmark. The fiscal balance (FB) results in a current account norm that is 0.7 percent of GDP lower than the average. The dependency ratio and population growth have offsetting effects—the latter reflecting the relatively rapid growth in prime-age savers—with a net upward effect on the current account balance equivalent to some 1.8 percent of GDP. India’s relatively high dependence on imported oil reduces its current account norm by about ½ percent of GDP. Regarding India’s stage of development, per capita growth and its relative income reduce the current account norm by some 2½ percent of GDP. The overall benchmark is thus about 3 percent of GDP, compared with an underlying current account of 2.7 percent of GDP.

22. The implied adjustment in the exchange rate is then calculated as the adjustment needed to reconcile the underlying and benchmark current accounts. The estimates are imprecise, in part due to the limitations of the underlying theory and associated empirical model, and in part due to the sampling error in calculating trade elasticities. While the attendant uncertainty is difficult to calculate, it could potentially be large.9 Nevertheless, the macro balance approach—like other estimates of exchange rate valuation—provides a useful benchmark for the level of the real exchange rate.

23. For India, the methodology suggests an insignificant deviation of the real effective rupee from the level implied by the macroeconomic balance approach. In particular, the deviation is only 3 percent, which is modest both in terms of the overall variability of the rupee and the underlying precision of the approach. The other two CGER approaches (the “equilibrium real exchange rate approach” which takes into account terms of trade and other fundamentals, and the “external stability” approach that gauges the exchange rate level by examining the stability of the external debt path) similarly find small deviations from estimated equilibrium.

D. Exchange Rate Policy and Intervention

24. India’s de-facto and de-jure regime is a managed float. India maintains an official policy of a managed float with no target or preannounced path for the exchange rate. Consistent with this, there has been significant flexibility in the exchange rate, measured in several ways (see Figure I.13)

  • Historical volatility has increased over time, and now is at a level concomitant with that of other large emerging markets countries.

  • Similarly, implied volatility—calculated from the prices of foreign exchange options—has risen over time, at a variety of maturities.

  • The prices of so called “risk reversals” suggest that markets have a roughly symmetric point of view about the likely future path of the exchange rate.10

  • A measure of overall flexibility that includes changes in reserves shows a degree of flexibility roughly comparable to other large emerging markets countries.

Figure I.13.India: Exchange Rate Highlights

Sources: Data provided by the Indian authorities; CEIC Data Company Ltd; Bloomberg LP; and IMF staff estimates.

1/ The index is calculated by dividing the standard deviation of exchange rate movements by an index of exchange market pressure (the sum of exchange rate volatility and volatility in reserves, normalized by lagged base money). It takes values from zero to one. A lower value signifies relative inflexibility, with zero indicating a peg or a high commitment to inflation targeting.

2/ Positive number implies markets assigning a greater probability (or premium) to INR depreciating than to appreciating against U.S. dollar.

25. Amid sharply rising capital inflows, the Reserve Bank of India has intervened with the stated aim to smooth volatility in the exchange rate. During January–October 2007 (latest available data), intervention amounted to about $64 billion, consisting of purchases of foreign currency. This represents a sizeable increase over 2005 and 2006, when intervention through October registered around $15 billion. During 2007, intervention has been highly variable from month to month, ranging from under $2 billion (in August) to over $12 billion (in October) but has consisted entirely of purchases: the most recent sale of foreign currency was conducted in December 2006, in the amount of about $6.5 billion. The sizeable, one-directional intervention, and the large increase compared with prior years, raises a question about its effects on the rupee exchange rate.

26. Research at the Fund and at the Reserve Bank of India (RBI) finds only limited effects of the RBI’s foreign exchange intervention on the rupee. Research conducted for the Fund’s Asia Pacific Regional Economic Outlook (October 2007), using monthly data, found no evidence that intervention either reduced the level of the rupee against the dollar, or slowed the rate of appreciation, even after attempting to control for endogeneity (e.g., the fact that intervention responds to changes in the exchange rate, as well as vice versa) using two-stage least squares techniques (see the table). Indeed, the correlation between intervention (proxied by reserves changes) and the rupee/dollar rate implied that intervention was sometimes associated with a more appreciated rupee or a faster rate of rupee appreciation, consistent with intervention that “leaned against the wind.” There were some findings, albeit limited, that intervention was associated with lower exchange-rate volatility (consistently correct signs, but not consistently significant in a statistical sense). Similar findings were uncovered in a study conducted at the Reserve Bank of India (Pattnaik and Sahoo, 2003).

27. Four factors could underly the lack of evidence for an effect of intervention. First, persistent structural factors—such as a wide productivity differential in favor of India—may be driving the appreciation of the currency, limiting any effect of intervention beyond a short period. Second, it is possible that the sterilization necessitated by the risk of stoking inflation, lead to persistent interest differentials, consequent capital inflows, and thus limited observed effects on the currency.11 Third, market participants may perceive that intervention does not portend coming changes in monetary policy—e.g., that intervention does not necessarily signal a looser-than-otherwise future monetary stance—limiting the effect through the “signaling channel.” Finally, the effects on market liquidity may be modest, given the high level of turnover in India’s foreign exchange market (some $34 billion per day, according to the latest figures from the Bank for International Settlements).

E. Conclusions

28. The appreciation of the rupee appears to be largely an equilibrium phenomenon. Consistent with this, the rupee does not appear to be out of line with medium-term macroeconomic fundamentals. Export performance has remained favorable, underpinned by structural improvements in the export sector, and more broadly by strong productivity growth. While intervention has risen in tandem with swelling capital inflows, empirical evidence suggests that intervention has served to dampen the volatility in the rupee, rather than to influence the level or rate of change in the currency. Going forward, continued efforts to maintain strong productivity growth, while allowing due flexibility in the currency, would be the best way to cope with any pressures on competitiveness likely to arise from any emergent currency appreciation pressures.

Table I.1.Asia: Real Effective Exchange Rates and Export Performance(Annual percent change)
200420052006Period Average
REERExports 1/REERExports 1/REERExports 1/REERExports 1/
China-2.635.4-0.228.52.127.1-0.330.3
India1.731.94.229.9-1.321.21.527.7
Indonesia-4.817.3-1.319.717.133.03.723.3
Korea1.830.912.112.17.416.27.119.7
Malaysia-4.420.50.311.44.014.00.015.3
Philippines-3.29.57.03.911.114.05.09.1
Thailand-0.519.81.914.58.918.73.417.7
Source: DGCI&S, India; and IMF, Information Notice System and Direction of Trade Statistics.

Goods export

Source: DGCI&S, India; and IMF, Information Notice System and Direction of Trade Statistics.

Goods export

Table I.2.CA Norm Contributions
TotalFB/GDPDep.

Ratio
PopOilBal/GDPPer Capita GrowthRel.

Income
CA Lag
India-3.1-0.72-1.2-0.6-0.6-1.70.1
Table I.3.MB Approach
ElasticitiesCA/GDP normChange in REER from reference Period to Projection DateProjected Medium-term CA/GDP at Reference Period Exchange RateMultilaterally Consistent RER gap
(In percent, except for elasticity)
0.19-3.1-1.4-2.7-3
Table I.4.Correlation between Intervention and Rupee/Dollar Rate
2000–20072000–20022003–20042005–2007
Level1-0.21*0.46-0.35-0.22
Change2-0.31*-0.48*-0.45*-0.17
Volatility30.05-0.39*-0.11-0.09
* indicates significance at the 95 percent confidence level.Source: “Sterilized Intervention in Emerging Asia: Is It Effective?,” Chapter III in Asia Pacific Regional Economic Outlook, Asia and Pacific Department, International Monetary Fund, October 2007

A positive number indicates that intervention is associated with a more depreciated rupee.

A positive number indicates that intervention is associated with slower rupee appreciation.

A negative number indicates that intervention is associated with lower rupee volatility.

* indicates significance at the 95 percent confidence level.Source: “Sterilized Intervention in Emerging Asia: Is It Effective?,” Chapter III in Asia Pacific Regional Economic Outlook, Asia and Pacific Department, International Monetary Fund, October 2007

A positive number indicates that intervention is associated with a more depreciated rupee.

A positive number indicates that intervention is associated with slower rupee appreciation.

A negative number indicates that intervention is associated with lower rupee volatility.

References

Prepared by Hiroko Oura, Petia Topalova, Andrea Richter-Hume, and Charles Kramer.

Several countries—notably Japan and Korea—have been able to sustain rapid export growth over several decades despite significant real currency appreciation. The Balassa-Samuelson effect provides an explanation for this in some cases. As the relative productivity of labor in tradables rises, wages in this sector increase, in turn raising demand for nontradables. This drives up average prices and causes the real exchange rate to appreciate. While this effect has been found to hold empirically for Japan and other advanced countries, evidence of its relevance in emerging economies is weaker.

Industrial productivity data (April 2007 WEO) are used as a proxy for productivity in the tradables sector.

See Bosworth and Collins (2006).

Real exports were obtained by deflating the reported U.S. dollars values by the U.S. GDP deflator.

HS codes were matched to NIC 1987 using the concordance by Debroy and Santhanam (1992).

Thus, when India expands its exports of tea to a country that previously was not importing Indian tea for example, the number of exported varieties rises.

The figures are based upon the Fall 2007 CGER exercise. Kohli and Mohapatra (2007) present an analysis of exchange-rate trends over 1980–2002.

Kramer (1996) finds large confidence intervals for the equilibrium value of the Canadian dollar using a Monte Carlo approach to the macroeconomic balance concept.

The risk reversal reflects the relative prices of foreign currency puts and calls—e.g. options that pay off if the currency appreciates or depreciates respectively. If the prices are roughly the same, then market participants put roughly equal probability (or risk premium) on appreciation or depreciation.

The notion that unsterilized intervention would affect the exchange rate is uncontroversial, as it would normally give rise to interest-rate changes that would affect currency values.

Other Resources Citing This Publication