Annex I. Derivation of the CMS Decomposition
The decomposition of the change in exports (X1 -X0) is derived as follows:
An early statement of the CMS methodology can be found in see Richardson, D. 1971, “Constant-Market-Shares Analysis of Export Growth”, Journal of International Economics, 1, pp. 227-239. For alternative formulations and refinements of the methodology, see Milana, C. 1988, “Constant-Market-Shares Analysis and Index Number Theory”, European Economy, Vol. 4, pp. 453-478
Arghyrou, M. and E. Bazina, 2002, “Competitiveness and the External Trade Performance of Greece in the 1990s,” Public Policy Discussion Paper 02-06, Brunei Business School, Brunei University.
Blanchard, O., 2006a, “Current Account Deficits in Rich Countries,” Mundell-Fleming Lecture delivered at the IMF’s 7th Jacques Polak Annual Research Conference.
Dimelis, S., 2004, “Comparative Advantages of the Greek Economy: An Aggregative and Sectoral Analysis” Center of Planning and Economic Research, Studies 53.
Dimelis, S., 2004a, “Prospects for the Greek economy in the Post-Enlargement Era,” The World Economy, Vol. 27, No. 6, pp. 803–827.
Driver, R. L., and P. F. Westaway, 2004, “Concepts of Equilibrium Exchange Rates,” Bank of England Working Paper No. 248, London.
Isard, P. and H. Faruqee (1998), “Exchange Rate Assessment: Extension of the Macroeconomic Balance Approach”, IMF Occasional Paper 167
Isard, P., R. Kinckaid, M. Fetherstone and H. Faruqee (2001), “Methodology for Current Account and Exchange Rate Assessments” IMF Occasional Paper 209
MacDonald, R., 2000, “Concepts to calculate Equilibrium Exchaneg Rates: an Overview,” Economic Research Group of the Deutsche Bundesbank, Discussion Paper 3/00.
Sideris, D and N. Zonzilos (2005), “The Greek Model of the European System of Central Banks Multi-Country Model”, Bank of Greece Working Paper 20
Prepared by Anastassios Gagales (EUR).
Unless otherwise stated, all figures are based on the national accounts prior to the September 2006 revision owing to the lack of sufficiently long GDP series (the new estimates, which are still provisional, are available only since 2000). The revision raised GDP by about 26 percent and reduced correspondingly all ratios to GDP. Although underlying economic conditions are unaffected by the revision, external imbalances appear less acute than previously.
Cross country differences in export deflators need to be interpreted with caution because, in addition to price effects, they reflect changes in the commodity composition and quality of exports, and the degree of market power. These are welfare improving and do not imply competitiveness loss.
See also the 2006 Report of the Greek National Council for Competitiveness and Development.
Credit to households, the main beneficiaries of financial liberalization, has been growing at almost 30 percent annually and its “contribution” to nominal GDP growth (measured by the ratio of credit expansion over last year’s GDP) reached 8½ percent in 2005. Although indicative of very easy monetary conditions, this rapid credit expansion is not entirely exogenous as it incorporates the effect of higher growth (actual and prospective), lower real interest rates, etc. Moreover, its impact on aggregate demand has been blunted by the increase in financial asset holdings and the substitution of bank for non-bank credit.
See Isard and Faruqee (1998), Chapter V, for a detailed discussion of the methodology of calculating CAU. The present paper deviates in that it abstracts from lagged exchange rate effects.
The elasticity of the current account with respect to the exchange rate is given by (X/Y)ηX + (M/Y) (ηM − 1) where (X/Y) and (M/Y) denote the shares of exports and imports to nominal GDP and ηX and ηM the export and import elasticities with respect to the exchange rate. The calculations are based on MULTIMOD elasticities (ηX = 0.71, ηM = 0.92). The quarterly BoG component (Sideris and Zonzilos, 2005) reports similar long-run elasticites (ηX = 0.6, nM = 0.9).
A comprehensive discussion of the different approaches for estimating the equilibrium exchange rate can be found in Driver and Westaway (2004). See also MacDonald (2000, 2006), MacDonald and Stein (1999), and Williamson (1994).
The two cornerstones of the Balassa-Samuelson effect is that the price of tradables are the same at home and abroad and wages are set so as to ensure the competitiveness of the traded goods sector. The latter is assumed to set the tone for the level of wages at home.
n assessing the significance of the Balassa-Samuelson effect it is important to identify whether the productivity growth stems from efficiency gains or from the closure of unprofitable firms. Only the former can support a thriving export sector and raise productivity growth on a sustainable basis.
CAS is derived from the equation of motion of NFA under the simplifying assumption of zero valuation effects on NFA and zero transfers.
To some extent, the high net foreign liabilities reflects measurement problems in the numerator and the denominator. Official statistics are unlikely to fully capture real estate and financial assets that Greek residents accumulated abroad prior to financial liberalization; extensive tax evasion continues to contribute to the under-reporting of foreign assets; and the recent upward revision of the national accounts (which has not been used in these calculations) reduces the ratio of net foreign assets to GDP to 56 percent.
Nevertheless, a large rise in external indebtedness would bring increasingly more risky borrowers into the picture raising (at the margin) the cost of borrowing.
The calculations are based on MULTIMOD elasticities, as detailed in footnote 6.
Calculations are based on the UN Comtrade (Commodity Trade Statistics) database using annual data for the period 1992-2005. Commodities are grouped in nine categories (food, mineral fuels, chemicals, raw materials, textiles, metals, manufacturing items, transportation equipment, and other manufacturing goods) and export markets in 13 regions (Germany, Italy, UK, France, Spain, Cyprus, Rest of EU15, USA, Turkey, Bulgaria, Romania, Albania, and the rest of the world).
This negative competitiveness effect is consistent with econometric evidence, based on longer samples, which suggest a below unity elasticity of exports with respect to world demand (Zonzilos, 2006).
New national accounts data, which capture better the underground economy, revised the 2005 current account deficit from 8 percent to 6 percent and the IIP from 82 percent to 65 percent.
According to preliminary flow of funds data, the decumulation of assets reflects banks’ running down of their government securities holdings to finance more profitable credit expansion. Households’ net financial position has declined very little, as the increase in bank borrowing was largely offset by an increase in households’ financial assets.
The calculation assumes that productivity in Greece remains on its current trend and wage moderation continues in the rest of the euro area. In this scenario nominal wages take the brunt of the adjustment. Supportive structural policies can take off some pressure from wages. For example, if structural reforms and capital deepening raise productivity growth by 1 percentage point to3½ percent per year, the adjustment period will be shortened to 4¼ years.
The new level of steady-state consumption is lower than the current rate and roughly equal to productivity growth plus population growth.
See Blanchard (2006a) for a similar analysis based on an optimizing model. In related work, Krugman (1987) argues that a long period of low production may permanently lower productivity and emphasizes the role of external learning by doing.
The above calculations refer to average wages. Actual wage settlements should ideally be guided by productivity and demand conditions in each individual industry. Wage restraint can be effected through fiscal policy and, in particular, by showing restraint in setting the public sector wage bill. This is of particular importance in Greece, where the public sector is large and sets the tone in wage negotiations.
The calculation assumes that the import content of exports is 30 percent. An import content of 40 percent would necessitate a further acceleration of exports by 1½ percentage points.