6. The purpose of this paper is to examine the adequacy of Cyprus’ economic competitiveness. Although Cyprus’ current account balance has fluctuated widely in the past two decades (Figure 1), some elements associated with the sharp deterioration of the current account from 1993-95 to 1996-97 may justify a closer analysis of the country’s competitiveness and current account sustainability. This paper looks at the recent behavior of conventional indicators of competitiveness, estimates Cyprus’ equilibrium real exchange rate, and provides an econometric analysis of the main determinants of service exports. The overall conclusion is that the country’s current level of the real effective exchange rate is in line with savings-investment fundamentals, but that the recent unfavorable trends in public savings and relative unit labor costs should be reversed to avoid problems in the medium term.

Abstract

6. The purpose of this paper is to examine the adequacy of Cyprus’ economic competitiveness. Although Cyprus’ current account balance has fluctuated widely in the past two decades (Figure 1), some elements associated with the sharp deterioration of the current account from 1993-95 to 1996-97 may justify a closer analysis of the country’s competitiveness and current account sustainability. This paper looks at the recent behavior of conventional indicators of competitiveness, estimates Cyprus’ equilibrium real exchange rate, and provides an econometric analysis of the main determinants of service exports. The overall conclusion is that the country’s current level of the real effective exchange rate is in line with savings-investment fundamentals, but that the recent unfavorable trends in public savings and relative unit labor costs should be reversed to avoid problems in the medium term.

II. External Competitiveness in Cyprus1

A. Introduction

6. The purpose of this paper is to examine the adequacy of Cyprus’ economic competitiveness. Although Cyprus’ current account balance has fluctuated widely in the past two decades (Figure 1), some elements associated with the sharp deterioration of the current account from 1993-95 to 1996-97 may justify a closer analysis of the country’s competitiveness and current account sustainability. This paper looks at the recent behavior of conventional indicators of competitiveness, estimates Cyprus’ equilibrium real exchange rate, and provides an econometric analysis of the main determinants of service exports. The overall conclusion is that the country’s current level of the real effective exchange rate is in line with savings-investment fundamentals, but that the recent unfavorable trends in public savings and relative unit labor costs should be reversed to avoid problems in the medium term.

Figure 1.
Figure 1.

Cyprus: The Current Account Balance

(In percent of GDP)

Citation: IMF Staff Country Reports 1998, 098; 10.5089/9781451809794.002.A002

Source: The Central Bank of Cyprus.

7. Cyprus’ current account position has deteriorated significantly in the last few years, moving from an average balanced position in the period 1993-95 to an average deficit of 5 percent of GDP in 1996-97.2 This deterioration stemmed solely from the balance in goods and nonfactor services, as net transfers and investment income as a percentage of GDP remained virtually unchanged in the period. The worsening in the balance of goods and nonfactor services can be partially explained by the negative shocks that occurred in 1996-97, namely the severe drought that affected agricultural exports in 1996 and 1997 and the escalation of tensions between the Greek and Turkish Cypriots, which significantly reduced the number of tourists visiting Cyprus in 1996.3

8. Looking ahead, the unwinding of the temporary shocks described above should contribute to an improvement in the external accounts in the near term. There are other favorable indicators of Cyprus’ external position and current account sustainability as well:

  • the low external debt to GDP ratio, which decreased from 36 percent in 1990 to 25 percent in 1997. This path reflected mainly the government’s choice to substitute domestic for foreign borrowing while private foreign debt remained fairly stable (Figure 2).4

  • the low and declining level of interest payments on foreign debt, which dropped from 4.6 percent of GDP in 1993 to 2.8 percent in 1997. As a percentage of exports of goods and nonfactor services (GNFS), interest payments in 1997 were equivalent to about 6 percent (Figure 3).

  • the level of official reserves, which in 1997 was equivalent to 18 percent of GDP or 5½ months of imports (Figure 4).

Figure 2.
Figure 2.

Cyprus: External Debt

(In percent of GDP)

Citation: IMF Staff Country Reports 1998, 098; 10.5089/9781451809794.002.A002

Source: The Central Bank of Cyprus.
Figure 3.
Figure 3.

Cyprus: Interest Payments on External Debt

Citation: IMF Staff Country Reports 1998, 098; 10.5089/9781451809794.002.A002

Source: The Central Bank of Cyprus.
Figure 4.
Figure 4.

Cyprus: Official Reserves

(In millions of US dollars)

Citation: IMF Staff Country Reports 1998, 098; 10.5089/9781451809794.002.A002

Source: The Central Bank of Cyprus.

9. Despite these favorable elements, however, the deterioration in Cyprus’ external accounts could be viewed as potentially worrisome in light of the relative cyclical positions of Cyprus and its main trading partners, and because (i) it has also been associated with negative trends in competitiveness and export market shares observed since the early 1990s and (ii) it has been underlined by a severe drop in public savings.

10. Output growth in Cyprus slowed from 1993-95 to 1996-97, creating a negative output gap, while the opposite trend occurred in its tourism and trade partner countries (Figure 5). One would expect these cyclical movements to lead to an improvement in the current account balance. Instead, the worsening of the current account balance took place despite the favorable impact of the relative cyclical positions, suggesting a strong structural deterioration that can only partially be attributed to the shocks. In fact, the estimated cyclically adjusted current account balance shifted from an average surplus of 3.8 percent of GDP in 1993-95 to an average deficit of 5 percent of GDP in 1996-97 (Figure 6).5 Furthermore, as opposed to the actual current account balance, the cyclically adjusted balance continued to worsen in 1997.

Figure 5.
Figure 5.

Cyprus: Output Gap

(In percent of potential output)

Citation: IMF Staff Country Reports 1998, 098; 10.5089/9781451809794.002.A002

Sources: IMF, World Economic Outlook; and staff estimates.
Figure 6.
Figure 6.

Cyprus: Current Account

(In percent of GDP)

Citation: IMF Staff Country Reports 1998, 098; 10.5089/9781451809794.002.A002

Sources: Central Bank of Cyprus; IMF, World Economic Outlook; Direction of Trade Statistics; and staff estimates.

11. Contrary to the large current account deficits in the 1980s and early 1990s, which were associated with high investment rates, the recent current account deficits occurred in a context of declining investment that was more than offset by a sharp reduction in saving (Figure 7).6 The decline in investment is part of a long-term trend occurring since the early 1980s, which has been attributed to capital deepening and lower profitability.7 The decline in aggregate saving between 1993-95 and 1996-97, equivalent to about 7 percentage points of GDP, can be partially explained by private consumption smoothing and a decline in public savings of the order of 2¾ percentage points of GDP.

Figure 7.
Figure 7.

Cyprus: Savings and Investment

(In percent of GDP)

Citation: IMF Staff Country Reports 1998, 098; 10.5089/9781451809794.002.A002

Sources: IMF, World Economic Outlook; and staff calculations

12. Notwithstanding the aforementioned shocks, some unfavorable trends in Cyprus’ external position and competitiveness have also played an important role in the current account deterioration of the mid-1990s. Increases in labor costs have outpaced gains in productivity and export prices through most of the 1980s and 1990s, thus resulting in declining profitability in export activities (Figure 8). Relative unit labor costs (ULC) have increased steadily since the early 1990s, contributing to some loss in service export market share and an even sharper decline in the market share of domestic exports in world trade, especially in the EU (Figures 9 and 10).8 The lost dynamism of domestic exports of goods and services has been only partially offset by an increase in (lower value added) reexports, notably to Russia. As a result, total exports of GNFS trended downward, having averaged 53 percent of GDP in 1982-90 against 49 percent of GDP in 1990-97; combined with the recent rise in imports, this trend may raise questions about the adequacy of competitiveness of the economy as a whole. The next section looks further into this issue.

Figure 8.
Figure 8.

Cyprus: Export Profitability 1/

(Index; 1979=100)

Citation: IMF Staff Country Reports 1998, 098; 10.5089/9781451809794.002.A002

Sources: The Central Bank of Cyprus; and staff calculations.1/1/ Export Deflator/ULC
Figure 9.
Figure 9.

Cyprus: Competitive Indicators

(Index; 1988=100)

Citation: IMF Staff Country Reports 1998, 098; 10.5089/9781451809794.002.A002

Sources: IMF, International Financial Statistics; Central Bank of Cyprus; OECD; and staff calculations.1/ ULC in the whole economy vis-a-vis Cyprus’ main industrial partner countries (tourism weights).2/ Vis-a-vis Cyprus’ main industrialized trade partners.
Figure 10.
Figure 10.

Cyprus: Export Market Share

(Index; 1981=100)

Citation: IMF Staff Country Reports 1998, 098; 10.5089/9781451809794.002.A002

Sources: Central Bank of Cyprus; IMF, World Economic Outlook; and staff calculations

B. Estimating Cyprus’ Equilibrium Real Effective Exchange Rate

13. This section estimates Cyprus’ equilibrium real effective exchange rate based on a macroeconomic balance framework that focuses on the requirements for simultaneously achieving internal and external balance.9 This process involves estimating the current account that is likely to emerge in the medium term under prevailing exchange rates and with domestic and foreign outputs at their potential levels (the underlying current account) and comparing it with the “equilibrium” current account. The equilibrium current account here refers to the level of the current account that is compatible with the estimated savings/investment relation at full employment. When the underlying and equilibrium current account balances differ, it is assumed that the real exchange rate will adjust so as to move the underlying current account to the level of equilibrium. In the calculations reported below, it was assumed that the medium-term positions are reached over a three-year period.

14. The underlying current account is the cyclically adjusted current account that would result at current exchange rates if these rates were sustained for three years. It is calculated by adjusting the actual current account for existing output gaps in Cyprus and abroad and for the lagged effects of recent real exchange rate movements. The first step in performing this calculation was to estimate the long-run income and exchange rate elasticities of Cyprus’ exports of goods, exports of services, imports of goods, and imports of services.10 A dummy variable accounting for the 1974 invasion was included in the export equations. Results are shown in the table below, and all coefficients are significant at the 1 percent significance level:

Long-Run Export and Import Elasticities

article image
Source: Cyprus authorities; and Fund staff estimates.

The estimated elasticities for traded goods are broadly in line with the elasticities estimated by Ito and others (1996) for APEC countries in a similar stage of development. The fact that the elasticities of services exports are much higher reflects the large share of tourism exports in total service exports, and the high sensitivity of the tourism activity to price and income fluctuations.

15. Once the long-run elasticities were estimated, the next step was to follow the CGER assumption that the full impact of an exchange rate change on trade (and service) volumes is spread over three years, with 60 percent of the long-run effect occurring during the first year, a further 25 percent during the second year, and the remaining 15 percent in the third year. It is also assumed that the price of each country’s (aggregate) exports is unaffected by changes in its real exchange rate, while its import price immediately adjusts to reflect fully any changes in relative costs.

16. The output gaps of industrialized countries are IMF estimates used for the World Economic Outlook exercise, whereas the output gaps for Cyprus and developing partners were estimated as deviations from trend GDP obtained through Hodrick-Prescott filtering. Weights used in the averaging of output gaps and exchange rate movements correspond to each country’s share in Cyprus’ exports of goods and exports of services. These estimates and assumptions, together with the observed current account deficit of 4½ percent of GDP, lead to an estimated underlying current account deficit of 5 percent in 1997. This means that at potential output (both domestic and foreign) and after factoring in the lagged effects of recent exchange rate fluctuations, the current account deficit in 1997 would be equivalent to 5 percent of GDP.

17. The equilibrium, or structural, current account balance is derived from a model of the equilibrium saving-investment balance which relates investment and saving flows (and thus the current account) to their medium-term structural determinants, notably country-specific and global fiscal variables, demographic variables, and relative country size.11 The model was estimated with panel data from 21 OECD countries over the period 1971-93. Using the long-run coefficients obtained from this model, and a constant term that captures effects specific to Cyprus,12 the structural current account balance for Cyprus can be derived as

CA=0.65SUR0.20DEM+0.09YPCAP1.6,(1)

where:

article image

The equation captures the idea that high public savings and low population dependency ratio contribute to domestic savings and thus a higher current account balance, whereas low levels of income per capita tend to be associated with high returns to capital and thus high investment rates and a lower current account balance.

18. According to these parameters, Cyprus’ structural current account deficit in 1997 was equivalent to 4½ percent of GDP.13 This means that the model’s structural determinants of savings (the fiscal balance and the dependency ratio) and investment (per capita income) in Cyprus currently predict a need for supplemental foreign saving of the order of 4½ percent of GDP. For the sake of comparison, Cyprus’ actual current account deficit in 1985-95 averaged 3½ percent of GDP. Since the average fiscal deficit during that period was also lower than the current fiscal deficit, the estimated structural current account deficit in 1997 does not seem out of line with Cyprus’ experience.

19. Counterposing the estimated underlying and structural current account deficits, and using the exchange rate elasticities reported in the table on page 19, it can be concluded that the 1997 real effective exchange rate of the Cyprus pound is broadly in line with present savings and investment fundamentals. It must be emphasized, however, that this result (i) is a function of the current state of the public finances, and (ii) does not involve any judgement or projection regarding the long run sustainability of the fiscal and external positions.

20. The present levels of fiscal and current account deficits imply a rising debt to GDP ratio, a situation that is undesirable in the medium term and unsustainable in the long run. Having this concern in mind, an alternative CGER simulation was carried out considering a scenario in which a three-year fiscal adjustment lowers the structural public deficit to 3 percent of GDP, a level compatible with a stable public debt-to-GDP ratio according to staff calculations. This improvement in public savings would imply an estimated structural current account deficit of 3½ percent of GDP which, in turn, would yield an estimated overvaluation of the pound of about 4 percent. In other words, a real effective depreciation of 4 percent would be required to bring Cyprus’ (underlying) current account to its equilibrium balance if the structural fiscal deficit were to be reduced to 3 percent of GDP. This depreciation could take place without changes in the nominal exchange rate peg—for example, through a slowdown in wage and price inflation that would follow the envisaged contractionary fiscal policy. This adjustment path, however, is dependent on and reinforces the importance of reducing wage and price rigidities.

21. Another source of concern for the sustainability of Cyprus’ current account is the decrease in investment and in export profitability of recent years, which may reduce the country’s ability to generate foreign receipts in the near future. Cyprus’ growing dependency on tourism for foreign exchange receipts (which are more sensitive to price and demand swings) also increases the country’s external vulnerability and makes it even more important to address the factors hindering the sector’s development. The next section presents an analysis of the dynamism of service sector exports, with special emphasis on the extent to which labor costs affect the performance of exports of services.

C. Competitiveness in Exports of Services: A Vector Error Correction Analysis

22. Exports of services, notably tourism, have increased significantly since the late 1970s and have become Cyprus’ main source of foreign exchange earnings and main economic activity. While in 1980 only about 19 percent of total exports of GNFS came from tourism receipts and 28½ percent from exports of manufactures, in 1997 tourism brought in 41 percent of Cyprus’ exports of GNFS against only about 8½ percent obtained from exports of manufactures and 11½ percent from exports of domestic goods. Most of the decline in exports of domestic goods can actually be explained as a natural movement toward the exploitation of the island’s comparative advantage in tourism, as investment in the sector has grown much faster than investment in manufacturing (export) activities. The importance of the service sector as a foreign exchange source in Cyprus warrants the analysis below.

23. The performance of Cyprus’ exports of services is analyzed in this section based on the methodology used by Agénor (1995) to study the competitiveness of the French manufacturing industry. This methodology will provide an estimation of the relative importance of external demand, labor costs, and structural (nonprice) factors in the performance of Cyprus’ services exports during the last 27 years. It is particularly important to evaluate the impact of labor costs on the export of services, since tourism activity is very price elastic. The preliminary assessment in this analysis indicates that the stagnation of Cyprus’ tourism market shares may be related to the upturn in the island’s relative unit labor costs vis-à-vis its main tourism (and service) partners. The methodology consists of two steps: (i) estimating a cointegrating relationship between service exports and foreign demand, relative ULC, and an indicator of nonprice competitiveness; and (ii) estimating a vector error correction model for the variables described above, which is then used to compute variance decompositions and impulse response functions. Variance decompositions can be used to estimate the share of the variance of each variable that is explained by the other variables in the system, while impulse response functions show how innovations to one variable affect the behavior of the other variables through time.

24. The principal variables of the analysis are defined as follows, where all variables are annual values in natural logarithms:

article image
article image

The sample period is 1970-97. All series were tested for stationarity using augmented Dickey-Fuller tests. The results presented in the table below indicate that all series are nonstationary in levels but are stationary in first differences at the 5 percent significance level (for which the critical value is -2.98).

Augmented Dickey-Fuller Test Statistic

article image

25. The cointegrating model yielded the following long-run relationship:

XSERVt=1.92*PCGDPt0.18RULCSt+0.61TISt5.1+Et,(2)

where Et is the error term and Johansen tests indicated that the series are cointegrated. The error correction formulation corresponding to this model is estimated as

DXSERVt=0.30*DXSERVt1+0.64*DXSERVt20.82*PCGDPt10.77*PCGDPt20.14RULCSt10.21*RULCSt20.18*TISt10.19*TISt2+0.26*Dt0.40Et,(3)

where Et is the residual in equation (2).15 The model estimation has a coefficient of determination (R-square) of 0.77. The coefficient of E, is relatively large, indicating a rapid model convergence toward long-term values.

26. The model described in equations (2) and (3) yields the impulse response functions depicted in Figure 11 and the variance decomposition shown in the table below.16 As expected, external demand is the most powerful of the three explanatory variables in the short run (the current year), with relative ULC having little contemporaneous impact on the volume of services exported. After five years, however, the (negative) influence of increases in relative ULC peaks and explains about 20 percent of the variance of export volumes. In the long run, that is, after 10 years, external demand explains 31 percent and relative ULC are responsible for 18 percent of the variance in export volumes, while investment in the tourism sector explains about 12 percent.17 These results confirm the importance of cost factors in Cyprus’ export performance. Looking ahead, measures to contain the sustained growth in labor costs will be key to the further development of services exports.

Figure 11.
Figure 11.

Cyprus: Impulse Response Function for XSERV 1/

Citation: IMF Staff Country Reports 1998, 098; 10.5089/9781451809794.002.A002

Source: Staff calculations.1/ Response to one standard deviation innovations.

Variance Decomposition of XSERV

article image

D. Concluding Remarks

27. The improvement in the current account in 1997 interrupted a three-year trend of deterioration and helped alleviate the concern over Cyprus’ external position. CGER estimates indicate that Cyprus’ real exchange rate is in line with present savings-investment fundamentals. These fundamentals indicate the need for a high inflow of foreign savings, especially in view of the high fiscal deficit, and thus the appropriateness of the current exchange rate to generate such inflows (or such current account deficits). This analysis, therefore, suggests that the appropriate way to improve the current account to a position that is both sustainable in the long run and compatible with Cyprus’ development needs is to improve the fiscal balance, thus reducing the need for foreign savings and allowing for a lower real exchange rate. The effectiveness of the fiscal adjustment in enhancing Cyprus’ competitiveness within the prevailing exchange rate regime will in turn depend on the flexibility of domestic prices and wages, which highlights the importance of abolishing the existing wage indexation system.

28. The declines in investment and in the profitability of export activities may hinder the economy’s ability to generate foreign exchange in the medium term. The recent decline in service export market share, the stagnation of exports as a percentage of GDP, and the appreciation of the real effective exchange rate (especially the ULC-based REER) simultaneously reflect and signal trends that, if left unchecked, may result in a worsening of Cyprus’ external competitiveness. Together with expansionary fiscal and credit policies, these trends could not only increase the current account deficit but also reduce the flow of capital into the country, leading to a fast deterioration of Cyprus’ ability to service its external obligations. The degree of dependency on tourism receipts adds to the vulnerability of the external accounts, owing to the sensitivity of this activity to price, costs, and external demand fluctuations. These findings lend further support to the recommendations that the fiscal deterioration be reversed and the upward trend in ULC be arrested to maintain the viability of the fixed exchange rate in the medium term.

References

  • Agénor, Pierre-Richard, 1995, “Competitiveness and External Trade Performance of the French Manufacturing Industry,” IMF Working Paper No. 95/137 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Cuthbertson, Keith, Stephen Hall, and Mark Taylor, 1992, Applied Econometric Techniques (Michigan: University of Michigan Press).

  • Debelle, Guy, and Hamid Faruqee, 1996, “What Determines the Current Account? A Cross-Sectional and Panel Approach,” IMF Working Paper No. 96/58 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Greene, William, 1990, Econometric Analysis (New York: Macmillan Publishing Company).

  • Hamilton, James, 1994, Time Series Analysis (New Jersey: Princeton University Press).

  • International Monetary Fund, SM/97/252, October 6, 1997, A Methodology for Exchange Rate Assessments and its Application in Fund Surveillance over Major Industrial Countries.

    • Search Google Scholar
    • Export Citation
  • Ito, Takatoshi, Peter Isard, Steven Symansky, and Tamin Bayoumi, 1996, Exchange Rate Movements and Their Impact on Trade and Investment in the APEC Region, IMF Occasional Paper No. 145 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Lipschitz, Leslie, and Donogh McDonald, 1992, Real Exchange Rates and Competitiveness: a Clarification of Concepts, and Some Measurements for Europe, Empirica, Vol. 19.

    • Search Google Scholar
    • Export Citation
  • Milesi-Ferretti, Gian, and Assaf Razin, 1996, “Current Account Sustainability: Selected East Asian and Latin American Experiences,” IMF Working Paper No. 96/110 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Shirai, Sayuri, 1996, Investment in Cyprus, in IMF Staff Country Report No. 96/16 (Washington: International Monetary Fund).

1

Prepared by Claudio Paiva.

2

The current account deficit declined from 5.4 percent of GDP in 1996 to 4½ percent of GDP in 1997. The cyclically adjusted current account deficit, however, continued to rise, as will be explained in paragraph 5.

3

The direct impact of these shocks on the cumulative value of exports earnings is estimated at 3⅓ percent of GDP: the drop in tourism receipts from 1995 to 1996 was equivalent to 2¼ percentage points of GDP, and the drop in agriculture exports from 1995 to 1996-97 is estimated at 1.1 percentage points of GDP.

4

Notwithstanding the large issuance of public bonds in the European market in 1997, there was a marked decrease in the overall debt ratio. Public foreign debt as a percentage of GDP decreased from 23 percent in 1990 to 11½ percent in 1997.

5

The table on page 19 shows the income elasticities used to adjust the values of exports and imports of goods and services to the levels that would have prevailed were output to equal potential output for both Cyprus and its trading partners.

6

The average investment to GDP ratio fell from 20.9 percent in 1993-95 to 18.8 percent in 1996-97. The average saving to GDP ratio dropped from 20.9 percent to only 12.9 percent.

8

Cyprus has however benefited greatly from the opening of the Eastern European economies, which have become an important market for Cypriot exports, especially reexports.

9

This framework was developed by the IMF’s Coordinating Group on Exchange Rate Issues (CGER) and is regularly used to estimate equilibrium exchange rates for industrialized countries. See International Monetary Fund (1997).

10

The long-run elasticities were estimated using annual data for the period 1970-97. The estimation procedure utilized was suggested by Engle and Yoo and described in Cuthberston, Hall, and Taylor (1992). This same procedure was applied by Ito and others (1996) to estimate income and real exchange rate elasticities of trade for 15 countries of the APEC region.

12

Since Cyprus was not among the countries considered in the panel-data estimations, the country-specific constant term (which corresponds to the general constant plus a fixed-effect dummy variable) for Cyprus was taken as the one that minimizes the squared errors between the observed current account deficits and the deficits predicted by the model coefficients.

13

The structural fiscal deficit in 1997 was estimated at 4.4 percent of GDP, against an average of 1½ percent in industrialized countries; Cyprus’ dependency ratio was approximately 56 percent, against an industrial country average of 66 percent; and finally, Cyprus’ potential GDP per capita was 50 percent of that of the United States, against an industrial-country average of 84½ percent.

14

The index is weighted by the countries’ participation in Cyprus’ total tourist arrivals.

15

VEC models do not provide estimates of structural parameters but rather unconstrained estimates of the relationships among the variables in the system, so that individual coefficients should be interpreted with caution. For further details see Hamilton (1994) and Agénor (1995).

16

When estimating impulse response functions and variance decompositions, the order in which the variables are entered into the system affects the outcome of the estimation. Such ordering must follow a progression of exogeneity of the variables, with the “most exogenous” (or those less affected by contemporaneous values of the other variables) being listed first and so on. In our case, exogeneity suggests the ordering PCGDP, RULCS, XSERV, TIS.

17

The simplicity of the model, with only three exogenous variables, and the relatively small sample size are probably behind the fact that a significant share of the variance of XSERV is explained by its own lagged values. When the variables are entered into the system in a different order, with TIS being allowed to explain contemporary changes in export volumes, the variance decomposition shows a much lower degree of persistence in XSERV, and investments in the service sector become the main explanatory variable in the long run. Nonetheless, the relative importance of external demand and relative ULC is not affected by this different ordering.

Cyprus: Selected Issues
Author: International Monetary Fund