Information about Middle East Oriente Medio
Chapter

Chapter 4: Approaches to Assessing the Exchange Rate Level

Author(s):
International Monetary Fund
Published Date:
March 2007
Share
  • ShareShare
Information about Middle East Oriente Medio
Show Summary Details

Following a long period of multiple and unstable exchange rates in Iran, the exchange rate was successfully unified in March 2002. Since then, Iran has adopted a managed float exchange rate regime and eliminated most exchange restrictions for current account transactions, culminating with the acceptance in August 2004 of obligations under Article VIII, Sections 2, 3, and 4 of the IMF’s Articles of Agreement. Like many other developing countries with managed floats (Hinkle and Monteil, 1999; and Ishii, 2003), the Iranian authorities continue to pay close attention to the nominal exchange rate. This raises theoretical and practical questions about how to determine the appropriate level of the exchange rate and how to initiate corrections if misalignments are detected.

Finding indicators to assess the exchange rate level, let alone estimating an equilibrium REER, is a major challenge in the Iranian context. The difficulties stem from the large share of oil exports in trade and the various distortions created by government intervention in economic activity. Theoretical and empirical cross-country research provides guidance on how to integrate the terms of trade in the analysis of exchange rate level in commodity-producing countries, but these studies are less relevant for a country like Iran, because of the prevalence of price and interest rate controls and the government’s direct intervention in allocating resources to priority sectors and enterprises. Frequent changes to trade policy and exchange restrictions, as well as numerous tax exemptions and subsidies, add another layer of complexity.

This chapter suggests a range of indicators that could be used to assess the exchange rate level. However, because of the difficulties noted above, no single indicator is given prominence; the particular combination of these indicators that would be best for assessing the appropriate level of the exchange rate remains a matter of judgment.

The chapter begins with a brief review of the institutional setup of the foreign exchange market during 1993–2003, which provides background and identifies key stages in the evolution of the exchange rate regime. Econometric techniques are used to analyze developments in the REER, including testing a purchasing power parity hypothesis and, in a more refined framework, incorporating productivity trends, terms of trade developments, and monetary policy variables. This empirical investigation is complemented by a discussion of other possible means of assessing the exchange rate level, including growth performance of the tradables sector and comparison of the actual external current account balance against a “norm” value derived from cross-country studies. The chapter concludes by suggesting that the level of the exchange rate in 2003/04 does not appear to have been misaligned, judging from the good performance of the tradables sector and the relative proximity of the current account balance to its long-term “normative” value.

Exchange Rate Policy in Iran since 1993

Since 1993, the exchange rate regime in Iran has undergone numerous changes in three distinct periods (Figure 9; Sundararajan, Lazare, and Williams, 1999; and Celasun, 2003).

Figure 9.Imports, Market Exchange Rate, and Trade and Payment Restrictions, 1992/93–2003/04

Sources: Iranian authorities; IMF staff estimates.

The first period coincides with the first attempt at exchange rate unification, which lasted for nine months during March–December 1993. This effort failed mainly because of an expansionary fiscal policy and an unfavorable external environment. Following the exchange rate unification, large foreign exchange losses associated with subsidized imports of basic consumption commodities and payments on public and publicly guaranteed debt resulted in sizable quasi-fiscal losses for the central bank. The monetization of these losses led to an acceleration in money growth to 35 percent in 1993/94 from 25 percent in 1992/93. This, along with some limited capital account and trade liberalization measures, put pressure on the foreign exchange market. The central bank faced difficulties in meeting the growing demand for foreign exchange without a large exchange rate adjustment,1 as oil prices started to decline and access to trade credits was significantly curtailed.2 The disequilibrium in the foreign exchange market culminated in a freeze on the official exchange rate that was applied to foreign exchange purchases used for essential imports and debt service payments. At the same time, an officially recognized market-determined exchange rate––the dealers’ market rate—emerged (Figure 10).

Figure 10.Exchange Rate Developments, 1993–2002

(Rials per U.S. dollar)

Sources: IMF, Information Notice System; Iranian authorities.

1 Initial observation: January 1993.

2 Initial observation: December 1993; final observation: March 2002.

3 Initial observation: May 1994; final observation: March 2001.

During the second period (1994–March 2002), various combinations of multiple exchange rates existed (Table 14). Nontariff import restrictions were gradually tightened through 1997, including strict import licensing and positive import lists for each segment of the foreign exchange market. Also, export surrender and repatriation requirements were introduced. Despite tighter trade and payment restrictions, the parallel market nominal exchange rate continued to depreciate sharply, as macroeconomic policies remained highly inflationary.

Table 14.Chronology of Exchange Rate Developments, 1991–2003
Starting DaheExchange RatesRials per U.S. dollarTransactions Taking Place at the Rates
1/21/91Basic official rate70Oil exports, imports of essential goods, imports for priority projects, official debt service, and government-supported students.
Competitive rate600Imports of intermediate and capital goods not included above as well as related services.
Floating rateFloating—determined by banks based on free market rateNon-oil exports, imports not covered by the above official rates, and most service payments and receipts.
Free nonbank market rateMarket-determinedForeign exchange freely bought and sold.
3/21/93Floating rate (unified official rate)Managed floatAll transactions (previously contracted debt were serviced at pre-March 1993 exchange rates and imports of certain essential goods were effected at the previous basic official rate through a government subsidy account during 1993/94).
Free nonbank marketMarket-determinedForeign exchange freely bought and sold.
12/21/93Official rate1,750All transactions initially. Gradually, the eligible payments became limited to imports of essential goods; payments for essential services, and official debt service.
Free nonbank market rateFloatingAll other transactions, including non-oil exports, tourist receipts, payment for nonessential goods and services, and transfers.
5/4/94Official rate1,750Oil exports, imports of essential goods, payments for essential services, and official debt service.
Official export rateInitially Rls 50 below authorized dealers’ rate, soon fixed at Rls 2,345.Non-oil exports and imports on a positive list.
Authorized dealers’ market rateFloating (Rls 2,680 on 12/31/94; as depreciated as Rls 6,200 in 5/95).All other transactions, including non-oil exports, tourist receipts, payment for nonessential goods and services and transfers.
Parallel offshore market rateMarket-determined (Rls 3,000–3,100 on 12/31/94).Foreign exchange freely bought and sold.
5/20/95Official rate1,750Oil exports, imports of essential goods, imports for priority projects, payments for essential services, and debt service on priority projects.
Official export rate3,000All other transactions, including non-oil exports, service receipts, and imports and service payments not covered by the official rate.
7/11/97Official rate1,750Oil exports, imports of essential goods, imports for priority projects, payments for essential services, and debt service on priority projects.
Official export rate3,000All other transactions including non-oil exports, service receipts; and imports and service payments not covered by the official rate.
7/11/97Exchange rate for import certiticates4,600Import certificates have been allowed to be traded through the TSE—up to 100 percent for carpet exporters and up to 50 percent for other non-oil exporters.
1/98Conversion of export proceedsDetermined at the TSEAll export proceeds could be sold at the TSE.
1/99Tehran Stock Exchange (TSE)Market-determinedThe TSE rate is allowed to depreciate and the queuing in the TSE is abolished.
3/00Export exchange rate3,000Abolished.
3/02Exchange rate unificationMarket-determinedA unified interbank market established.
2003Unified rateMarket-determinedOpening-up of the capital account for some ouflows through the off-shore foreign exchange market.
Source: Based on information provided by the Central Bank of Iran.
Source: Based on information provided by the Central Bank of Iran.

After 1997, a number of trade restrictions were gradually relaxed, which further fueled parallel market exchange rate depreciation. By 1998/99, macroeconomic policies were tightened and the parallel market exchange rate, as well as an increasingly market-determined exchange rate in the Tehran Stock Exchange, stabilized.3 In 1999, all administrative controls on the TSE exchange rate were lifted, laying the basis for a smooth transition toward a unified exchange rate system. Despite these liberalization efforts, stringent licensing requirements and a positive list for imports remained in place until March 2002. The parallel market for foreign exchange remained active up to the March 2002 unification inside the country and offshore, even though the premium over the TSE rate was on a declining trend.

The third period started with the March 2002 exchange rate unification and a substantial reduction in import restrictions. During 2003, some capital account liberalization measures were undertaken as well (Chapter 2). The authorities opted for a managed float exchange rate regime, taking into account the nature and magnitude of the shocks the economy was likely to face, as well as the structural characteristics of its goods, labor, and financial markets. Since Iran has been subject to both real and nominal shocks in the context of significant nominal rigidities, a managed float exchange rate regime appeared appropriate, because it was likely to reduce the output costs of adjustment to shocks.4 A low degree of dollarization of the financial system (Chapter 2) was another important consideration for adopting an exchange rate regime with some degree of flexibility.

The 2002 unification has been successful, as evidenced by the stability of the nominal exchange rate and the significant decline of the premium in the parallel exchange market. The success is mainly due to the following factors. First, Iran initiated the exchange regime change from a position of strength, benefiting from high oil prices, renewed reform efforts, and a few years of experience with a market-determined exchange rate in the TSE market, which had operated in the context of a dual exchange rate system. Second, the Iranian authorities were able to minimize the pass-through to prices of the depreciation of the official exchange rate, relying on resources previously accumulated in the Oil Stabilization Fund (OSF) and high oil revenue to subsidize basic consumer goods. As a result, following the exchange rate unification and the move to a managed float, the exchange rate quoted in the interbank market remained relatively stable, which helped bolster business confidence and ensure a smooth transition to the new system.

After the fall of 2002, the authorities initiated a gradual depreciation of the nominal effective exchange rate vis-à-vis a basket of currencies, while reverting to an expansionary policy mix. At the same time, a significant reduction of nontariff import barriers and some liberalization of capital outflows (Chapter 2) helped, to some extent, contain pressure on the nominal exchange rate to appreciate against the background of increasing government sales of oil-related foreign exchange to the market to boost expenditure. Despite these measures, double-digit inflation persisted.

REER-Based Measures of Competitiveness

The analysis in this chapter uses exchange rates from the IMF’s Information Notice System, which splices data on various representative market-related exchange rates and other exchange rate series provided by the central bank.5 The overly appreciated official exchange rate fixed at Rls 1,750 per US$1 during 1993/94–2001/02 is not used, because it was unrelated to fundamentals and mainly served redistribution purposes, which, in practice, could have been achieved through explicit taxes and subsidies (Mullin, 1993). The official rate mainly applied to energy-related exports and subsidized imports (whose pricing and distribution were subject to tight government oversight) and to debt service payments on public and publicly guaranteed debt. To calculate the REER, CPI inflation differentials with trading partner countries are used. Weights are derived from the composition of trade (REER1). To check the robustness of results to changes in assumptions on trade weights, an REER using weights from the non-oil trade composition is also calculated (REER2). Excluding oil trade from the REER calculations reduces the relative weights of the euro area and Japan—the major destinations of oil exports–and provides a better indicator of competitiveness of the non-oil economy in Iran.

Relative Price Indicators of Exchange Rate Level

The REER and the PPP

Purchasing power parity is a possible yardstick for exchange rate assessment. It is based on the law of one price (Dornbusch, 1987), which, under the absolute PPP definition, stipulates that every good should have the same price once converted into a common currency unit. If the law of one price holds, the values of identical baskets of goods should be the same across countries once they are converted into a common currency unit.

The most widely used methodology to confirm or reject the PPP is based on the analysis of the time-series properties of the REER, which is assumed to measure changes in price level differences between a country and its trading partners (Rogoff, 1995). If the REER series is stationary (that is, does not have a stochastic trend), the PPP should hold. Even if the REER series is not stationary, the speed of convergence of the REER toward its equilibrium should be fast enough (say, below one year) for the PPP to hold. Slow convergence would imply that disequilibrium could persist in the medium and long term (Rogoff, 1995). This version of tests is consistent with a relative PPP formulation, which accounts for changes in the CPI rather than changes in the prices for specific goods and services.

We apply this methodology to Iran’s REER series. Both REER series show a clear cyclical behavior with an upward trend (Figure 11), and the visual observations are confirmed by statistical tests. Both are nonstationary I(1) processes based on Augmented Dickey-Fuller (ADF) test statistics.6 As a result, a no-PPP hypothesis cannot be rejected in the case of Iran. This evidence is similar to the findings for other commodity producers (Cashin, Céspedes, and Sahay, 2002; and Chen and Rogoff, 2002).

Figure 11.Real Effective Exchange Rate Indices, 1993–2004

Sources: International Financial Statistics; IMF staff estimates.

It may be that these tests are not powerful enough to reject the null hypothesis of nonstationarity, and the series could actually be stationary (Kwiatkowski and others, 1992). If one supposes that the REER is stationary, given the short length of the series, the next step would be to determine the speed of REER adjustment to a unit shock. For this purpose, an AR(1) ordinary least squares (OLS) regression was run to estimate the half-life of a unit shock or impulse (HLS). The estimated autoregressive coefficient α is equal to 0.98 for both series; thus, the HLS7 was equal to 35 months. This length is similar to those of most commodity-based currencies (Cashin, Céspedes, and Sahay, 2002). As mentioned, a three-year half-life for adjustment is difficult to reconcile with the traditional PPP formulation, which allows for only short-term deviations.

The results of the tests are not consistent with the PPP theory conjecturing that the REER should be a mean-reverting I(0) process with a relatively rapid pace of adjustment. Although the PPP theory is rejected for Iran, assessing the REER level with respect to its long-term mean still provides a point of departure for further analysis. By March 2004, REER1 and REER2 were above their long-term means by 22 and 29 percent, respectively. The next step is to assess whether such a real appreciation is justified on the basis of additional considerations.

The REER and Real Long-Term Fundamentals

There is a strand of theoretical literature that explains drifting REER in developing countries by productivity differentials8 and terms of trade (De Gregorio and Wolf, 1994; and Obstfeld and Rogoff, 1996).9 Drawing on such theoretical models, empirical literature usually looks for a long-term relationship among REER, productivity differentials, and terms of trade. Based on co-integration tests (see Rogoff, 1995; Chen and Rogoff, 2002; and Cashin, Céspedes, and Sahay, 2002), there is empirical evidence that the REER in some commodity-producer countries is co-integrated with measures of relative productivity and/or terms of trade.

In this empirical study on Iran, productivity differentials are omitted from the co-integration analysis for two reasons. First, available measures of productivity differentials with trading partners are not fully compatible with the theoretical formulations presented above and thus cannot be included in tests. This said, these indirect measures point to a lack of convincing evidence of higher productivity growth in Iran relative to trading partners. In fact, labor productivity in Iran grew at about 2 percent per year on average during 1993–2003, which is close to the weighted average productivity growth in trading partner countries (Appendix 2). Low, or even negative, TFP growth during 1993–2003 (Chapter 1) corroborates the results presented in Appendix 2. Second, there is no reliable high-frequency measure of relative productivity for Iran. Indeed, using annual data instead of monthly series will preclude any meaningful co-integration analysis.10

A real price of oil was used as a proxy for terms of trade, in line with similar research on other commodity producers (Cashin, Céspedes, and Sahay, 2002). The index of the real oil price was calculated using the U.K. Brent spot price deflated by the manufactured exports unit price index for developed countries.11

A visual inspection of REER1, REER2, and the real oil price series reveals two features. First, there is no clear pattern of co-movement of either REER series with the real oil price, in particular during 1997–98 (Figure 12). Second, volatility of both REER series appears to be lower than that of the real oil price.

Figure 12.REER, M2, and Real Oil Price, 1993–2003

(Log 1999/2000 = 1)

Sources: Iranian authorities; IMF staff estimates.

These observations are confirmed by econometric tests. There is no co-integration between either REER1 or REER2 and the real oil price,12 implying that the latter cannot explain the long-term REER dynamics. The standard deviation of REER1 and REER2 is 0.2, and that of the real oil price is 0.3.13 This finding is surprising, because in most countries, including commodity producers, the REER is more volatile than fundamentals (MacDonald, 1999). This suggests that in Iran the market-related exchange rate was “managed” through intervention in the foreign exchange market and trade and payment restrictions, because the REER was less volatile than economic fundamentals.

The REER and Monetary Variables

The previous subsection shows that there is no statistically proven co-movement between the REER and the terms of trade. Thus, other factors that have a long-term impact on the REER should be considered. Monetary variables are potential candidates. In most theoretical settings, money has some short-term impact on real variables in open economy models, but it is neutral in the long run. But given mixed evidence behind the long-term neutrality of money, Obstfeld and Rogoff (1996) developed a model demonstrating that monetary shocks have short-term real effects due to sticky prices, and that they may also lead to temporary current account imbalances that have long-run wealth effects on the real exchange rate. However, they show that “the long-run effects must be smaller than the short-run effects” (page 681). Given these considerations, M2 was included in the co-integration relationship along with the terms of trade to explain the long-term dynamics of the REER in Iran. However, because in Iran fiscal policy directly affects monetary conditions (see Chapter 3), fluctuations of money supply will be a measure of monetary shocks and indirectly a proxy for the fiscal policy stance.14

Separate co-integration tests were conducted for REER1 and REER2. Other variables in both tests included the real oil price (LRPOIL) and seasonally adjusted M2 (LM2).15 A model with one lag and a constant in the co-integration vector was selected in both cases.16 The Johansen trace and eigenvalue tests indicate the existence of at least one co-integrating vector in both cases. The estimated co-integrating vector β is presented in equations (4) and (5):

The next step is to test whether long-term elasticity coefficients in β are different from zero. A restriction on long-term elasticity of money being equal to zero is rejected.17 Thus, according to equations (4) and (5), higher money supply causes the REER to appreciate in the long run. As noted in Chapter 3, Iranian monetary expansions (contractions) broadly coincided with fiscal expansions (contractions), thus making REER dynamics indirectly related to the fiscal policy stance through money supply. This finding suggests that the fiscal impulse accommodated by monetary policy easing initiated in 2000 has contributed to the recent REER appreciation.

The long-term impact of the real oil price on either REER1 or REER2 was not supported by the findings.18 Thus, the issue of an equilibrium REER value consistent with the real-side long-term fundamentals is unresolved. There is no empirical support for the hypothesis that the substantial real appreciation registered by March 2004 relative to the 10-year mean also stems from the positive terms of trade shock that occurred at that time, partly due to the short data span and the various forms of government intervention.

The REER and the Tradables Sector

Given the limitations of the econometric techniques, a qualitative assessment of competitiveness is all the more important. Growth of non-oil exports and output performance of import-competing sectors provide good gauges of the current state of competitiveness. On these measures of competitiveness, Iran has performed well during 1999/2000–2003/04. Real value added in the manufacturing sector and non-oil exports grew at double-digit rates in many years (Figure 13).

Figure 13.Real Annual Growth Rates, 1993/94–2003/04

(In percent)

Sources: Iranian authorities; IMF staff estimates.

It is difficult to reconcile such a good performance in the tradables sector with the appreciation in the REER. Since there is no clear evidence of strong improvement in productivity differentials, the explanation is likely to be related to the operating environment of the tradables sector. The magnitude of pricing advantages stemming from trade barriers and the cost savings resulting from subsidies are likely to be much larger than the fluctuations in the REER. For example, imports of cars were previously prohibited and then were subject to a 150 percent import tariff. Tariffs on other imports that compete with domestic products are still high, despite recent efforts toward trade liberalization. More important, exporters, as well as import-competing industries and agriculture, continue to benefit from tax exemptions, explicit subsidies, and negative real rates of charge on banking loans as well as heavily subsidized domestic energy prices.19 The latter factor is particularly important for petrochemical exports. Finally, easier access to imports of capital and intermediate goods, as well as a reduction in tariffs on these categories, has recently improved competitiveness in the tradables sector.

Forward-Looking Measures of the REER and the Current Account Norm

Since the exchange rate plays a key role in shaping external balances, the exchange rate level should be assessed in the context of an analysis of the savings-investment balance as well. The latter is usually discussed in economic literature from a forward-looking perspective (Obstfeld and Rogoff, 1996), and thus, the appropriate exchange rate level should also be consistent with the notion of an equilibrium intertemporal allocation of resources. In simple terms, if the observed current account balance is believed to be close to its long-term equilibrium level, the exchange rate can be considered to be close to equilibrium.

This chapter uses a normative approach to determine the appropriate (“equilibrium”) level of the current account balance. Chinn and Prasad (2003) conducted an empirical study that “provides an indication of the levels of current accounts that may be considered ‘normal’ for a country, based on a number of its macroeconomic attributes” (page 48). In panel regressions for developing countries, excluding Africa, Chinn and Prasad (2003) determined the following macroeconomic attributes that affect the level of the current account balance: the ratio of the government fiscal balance to GDP, the ratio of net foreign assets to GDP, demographic dependency ratios,20 terms of trade volatility, financial deepening, and relative PPP-adjusted incomes to the United States.21 A norm for the current account balance in Iran was derived using the results of a number of panel regressions presented by Chinn and Prasad. Depending on the specification, the current account balance norm estimates for Iran range from a small deficit to a surplus of 1½ percent of GDP for 2003/04 (Table 15). Because the estimated norm for the current account balance is reasonably close to the actual realization, the REER level appears broadly consistent with the “normal” level of the saving-investment balance, conditional on the fiscal stance in 2003/04 and other parameters.

Table 15.Normative Current Account Balance, 2004
Panel Regression1Panel Regression2
Five-year averageRegression coefficient2003/04Regression coefficient
Government balance30.0150.259–0.0020.167
NFA/GDP30.0780.0390.0980.036
Dependency ratio40.704–0.055
M3/GDP30.4960.037
Standard deviation of terms of trade0.1660.034
Relative income50.163–0.049
Relative income squared50.0270.114
Current account balance (–1)33.060.536
Normative current account balance60.01.6
Memorandum item:
Current account balance in 2003/0461.51.5
Sources: Chinn and Prasad (2003); IMF staff estimates.

Panel regression for developing countries, excluding Africa, OLS specification with time effects.

Panel regression for emerging markets, excluding Africa, for annual data, OLS specification with time effects.

As a fraction of GDP.

Ratio of 0–14 year-olds to 15–64 year-olds.

Ratio to the U.S. per capita income in purchasing power parity (PPP)-adjusted terms.

Ten percent of GDP.

Sources: Chinn and Prasad (2003); IMF staff estimates.

Panel regression for developing countries, excluding Africa, OLS specification with time effects.

Panel regression for emerging markets, excluding Africa, for annual data, OLS specification with time effects.

As a fraction of GDP.

Ratio of 0–14 year-olds to 15–64 year-olds.

Ratio to the U.S. per capita income in purchasing power parity (PPP)-adjusted terms.

Ten percent of GDP.

Policy Implications of Findings

The previous sections have shown the difficulty of determining an equilibrium REER value based on real and monetary variables. However, other elements of the analysis do not point to a large misalignment of the exchange rate. Because monetary and fiscal policies are important determinants of long-term REER dynamics, fiscal and monetary tightening, in the context of limited capital mobility, could help reduce the speed of appreciation of the underlying unobserved equilibrium value of the REER in the short term, which may also stem from the recent increases in oil prices. This policy mix will contribute to faster growth of the non-oil sector by containing REER appreciation.

Conversely, trying to engineer a nominal depreciation of the effective exchange rate while pursuing an expansionary fiscal policy might lead to a further appreciation of the unobserved equilibrium value of the REER through an increase in inflation and even a short-term real exchange rate appreciation that exceeds this value.

Looking beyond the short term, such factors as oil prices, reform-driven productivity gains, progress in trade liberalization, subsidy reforms, and an eventual full capital account liberalization could have a significant impact on the equilibrium REER. However, given the uncertainties surrounding future developments in these fundamental factors and their impact on the equilibrium REER, the task of exchange rate management would become even more challenging. Thus, increased flexibility in exchange rate management might be needed, which would give the market an increasingly important role in determining the exchange rate. The central bank would have to tolerate a greater degree of two-way nominal exchange rate movements, while focusing on achieving its monetary policy objectives. The benefits of greater exchange rate flexibility can only be reaped if the monetary policy framework is substantially strengthened (Chapter 3).

Conclusion

A judgment on the appropriate level of the REER in Iran can be based on a combination of considerations. The PPP does not appear to be a useful yardstick for exchange rate management. More refined econometric analysis does not appear to yield significant results. Although monetary policy and the underlying fiscal policy stance appear to have a long-term impact on the REER, the terms of trade do not appear to explain the long-term evolution of the market-related REER. Thus, the extent to which recent REER appreciation is justified by terms of trade developments cannot be inferred from the data sample.

Given the limitations of the econometric analysis, the REER level was analyzed in light of the performance of the tradables sector and the current account norm estimated for Iran based on cross-country panel data. Owing to good performance of the tradables sector and the proximity of the current account balance to the “norm,” the exchange rate level does not appear to have been misaligned. This conclusion is subject to important caveats regarding various distortions that have helped improve the performance of Iran’s tradables sector, but at the expense of efficient allocation of resources, which might undermine future growth performance.

The short- and medium-term challenge facing the authorities is to avoid a sharp real appreciation of the unobserved equilibrium level of the REER, which could be triggered by a combination of high oil prices and expansionary policies. In the longer term, greater flexibility in exchange rate management is needed in the face of uncertainties related to the external environment and the effects of important structural transformations envisaged in the authorities’ reform agenda.

Note: Prepared by Vitali Kramarenko.
1Gross official reserves fell below $3 billion, or 1.4 months of next year imports, at end-March 1994.
2Net inflows under letters of credit of $2.5 billion in 1992/93 were followed by net outflows of $5 billion in 1993/94.
3Beginning in 1997, a growing number of foreign exchange transactions were allowed to take place at the TSE exchange rate. Foreign exchange was not traded directly. Rather, certificates of export proceeds, which could be used for authorized imports, were allowed to be traded in the TSE at increasingly market-determined exchange rates.
4See Celasun (2003) and Jbili and Kramarenko (2003) for detailed analyses and other factors in the choice of exchange rate regime in Iran.
5The nominal effective exchange rate (NEER) for Iran was calculated using data on the unified exchange rate during 1993, the free market dealers’ rate (1994–mid-1995), and, after the closure of the dealers’ market, the parallel market exchange rate (mid-1995–97), the TSE rate during 1998–2002, and the unified interbank rates thereafter.
6ADF statistics are –1.24 (0.65) and –1.23 (0.66) for REER1 and REER2 in logs, respectively, with p-values in brackets.
7HLS = abs (log ½ /log (α)).
8Defined as the sum of sectoral productivity differentials (tradables and nontradables) across countries.
9In these models, improvements in terms of trade mainly affect the REER through an increase in nontradables prices, which result from higher wages in this sector following higher wages in the commodity-exporting sector in the aftermath of higher export prices in the presence of labor mobility. In Iran, most oil export earnings accrue to the budget. Thus, appreciation pressures following higher oil prices will stem mainly from expectations and improved business confidence if the underlying fiscal stance does not change.
10In many studies, relative productivities are measured indirectly as differentials between tradables and nontradables price indices, assuming a competitive equilibrium framework and market pricing mechanisms (Alberola and others, 1999). Such an approach is not relevant for Iran, because relative prices are not likely to reflect relative productivities, given the extensive price controls, widespread subsidies, and restrictive trade policy in effect for most of the period under review.
11On the basis of ADF test statistics with p-values in brackets–– -1.17 (0.68)––the null hypothesis of nonstationarity of the real oil price index for 1993–2004 cannot be rejected. This finding is in line with other studies that have used longer time series and more sophisticated techniques (Cashin, Liang, and McDermott, 2000).
12The Aikaike Information Criterion and Schwartz Criterion indicate that one lag is appropriate. There is no co-integration for a system with one lag. Tests were also performed for systems with up to eight lags with no evidence of co-integration.
13Brown-Forsythe test (with p-values in brackets)––52.7 (0.0) and 35.7 (0.0)––confirms that the variances of REER1 and REER2, respectively, are not equal to those of the real oil price.
14Using fiscal variables directly is difficult, owing to a lack of reliable high-frequency fiscal data.
15All variables are in logarithms. Seasonal adjustment was made using the U.S. Census X12 package (multiplicative adjustment).
16Based on Aikaike and Schwartz information criteria.
17The restrictions are rejected at zero percent (χ2 = 20.0) and zero percent (χ2 = 22.0) confidence levels for equations (4) and (5), respectively.
18The restrictions cannot be rejected at 34 percent (χ2 = 0.9) and 25 percent (χ2 = 1.3) confidence levels for equations (4) and (5), respectively.
19For further details, see Chapters 1, 2, and 5.
20The ratio of the population younger than 15 years old to the working-age population (15–64 years old).
21A dummy for oil-producing countries was insignificant in the panel regressions.

    Other Resources Citing This Publication