Pass-through in dollarized countries: should Ecuador abandon the US dollar?

Abstract

In this article, we examine the convenience of dollarization for Ecuador today. As Ecuador is strongly integrated financially and commercially with the United States, the exchange rate pass-through should be zero. However, we sustain that rising rates of imports from trade partners other than the United States and subsequent real effective exchange rate depreciations are causing the pass-through to move away from zero. Here, in the framework of the Vector Error Correction Model, we analyse the impulse response function and variance decomposition of the inflation variable. We show that the developing economy of Ecuador is importing inflation from its main trading partners, most of them emerging countries with appreciated currencies. We argue that if Ecuador recovered both its monetary and exchange rate instruments, it would be able to fight against inflation. We believe such an analysis could be extended to other countries with pegged exchange rate regimes.

Keywords:

• pass-through
• shocks
• dollarized countries
• structural VECM

JEL Classification:

• F31
• E31

View correction statement:

Erratum

Acknowledgements

This work was supported by the Government of Spain and FEDER under grant number ECO2010-21787-C03-01. The authors wish to thank Andrea Cipollini, Antonio Ribba and Barbara Pistoresi for very useful comments and suggestions.

Notes

1. Even the European Monetary Union, the benchmark for economies undertaking similar projects, has been questioned in terms of a deficient political and fiscal union (Issing, 2011).

2. Reinhart and Rogoff (2004) used market-determined exchange rates (from dual/parallel markets) and found 14 categories of exchange rate regimes, ranging from no separate legal tender or a strict peg to a dysfunctional ‘freely falling’ or ‘hyperfloat’.

3. Ecuador is a member of major Latin American economic organizations including UNASUR (www.uniondenacionessuramericanas.com), CAN (www.comunidadandina.org) and ALBA (www.alianzabolivariana.org), which in 2007 created the Bank of the South – a credit institution similar to the World Bank, and it is soon to join MERCOSUR (MERCOSUR/CMC/DEC. Nº38/11). The aim of these organizations is to create a South American Free Trade Area, using a new currency (the sucre), which was first used in 2010 as a virtual currency in at least two transactions between Ecuador and Venezuela. Ecuador is also diversifying its trade partners, with Asian countries being its leading exporters in 2010 (www.icex.es).

4. Reported by the World Trade Organization in 2011 Press Releases (PRESS/628).

5. This new system replaced the USD, which had served as the sole anchor currency for approximately 10 years, with a basket of currencies that was weighted to account for bilateral trade volume and bilateral investment.

6. Frankel (2005) in his article ‘Contractionary currency crashes in developing countries’ describes how depreciations of the national currency cause contractionary effects rather than an expansion in economies highly indebted in dollars.

7. With business cycles negatively correlated, the FC increases real volatility and the TC reduces it. If the anchor country is hit by a positive shock, two simultaneous processes will take place: (1) the anchor country will increase imports from the pegged country, positively affecting the GDP of this country through the TC, but (2) since the anchor country could issue a restrictive monetary policy in order to avoid over-heating, the increase in the interest rate will negatively affect the pegged country through the FC.

8. See Alesina and Barro (2001), De Nicoló et al. (2005) and Berg and Borensztein, E. (2000).

9. Peru, Uruguay and Bolivia operate a high degree of financial dollarization (a variety of partial dollarization) in which foreign assets exceed domestic assets but where each country maintains its own currency. Haiti operates a system of semi-official dollarization in which the foreign currency is legal tender, but where it plays a secondary role to the domestic currency for paying taxes and wages. Finally, only three countries are fully dollarized, namely Ecuador, El Salvador and Panama.

10. A conclusion corroborated by Mann (1986) and Taylor (2000).

11. McCarthy (2000) provides empirical evidence in confirmation of these hypotheses associating both exchange rate and GDP volatility with a lower exchange rate pass-through to domestic inflation, although these relationships were only strong at short horizons.

12. See Eichengreen and Hausmann (1999).

13. See A1 in Appendix A for details of data sources.

14. The methodology for calculating the real effective exchange rate is outlined in Rodriguez (1999). The countries included are the United States, Japan, Colombia, Germany, Italy, Spain, Brazil, Chile, Mexico, Venezuela, France, the United Kingdom, Peru, Belgium, Argentina, Netherlands, Panama and South Korea, which account for about 89% of Ecuador’s total trade.

15. See Hwang and Wu (2011)  for China, Wilson (2006) for Japan and Ma (1998) for Colombia, three of Ecuador’s leading trade partners.

16. They assume foreign price shocks to be equivalent to nominal exchange rate shocks because when the latter changes persistently without changes to either the real exchange rate or domestic prices, the change is only recorded in foreign prices and the nominal exchange rate.

17. The higher real income resulting from a boom leads to extra spending on services, which in turn raises their price (i.e. causes a real exchange rate appreciation, defined as the relative price of nontraded to traded goods), where the boom is experienced in the extractive sector, and it is the traditional manufacturing sector that is placed under pressure (Corden and Neary, 1982).

18. See the interview in http://internacional.elpais.com/internacional/2011/02/24/actualidad/1298502020_850215.html

19. See Lütkepohl (2004).

20. In a Monte Carlo simulation study, Lanne and Lütkepohl (2002) show that LLS tests, which estimate the deterministic term by a GLS procedure under the unit root null hypothesis, enable remarkable gains in size and power properties and perform best in comparison to those tests which accommodate a deterministic level shift by estimating the deterministic term by OLS procedures.

21. See Breitung et al. (2004).

22. The econometric analysis was implemented using JmulTi 4 software (www.jmulti.de).

23. If we had not obtained cointegration without the inclusion of dummies so as to take the structural breaks into account, then we would have included them, but it proved unnecessary because the structural breaks coincided in more than one variable. It is supposed that the cointegration relation absorbed these structural breaks. See Juselius (2006).

24. We reach the same conclusion with the Johansen Trace test (Johansen et al. 2000). In the test we specified 1 lag for the variables in levels, two level shifts (2001:M2 and 2009:M1) unrestricted in the model, but seasonal dummies, intercept and trend restricted in the model. We estimate our VECM with the Johansen reduced rank, keeping this structure.